Topic Matches: [#513 - 910] / 910

513. In a simplified way, the hypertextual technology has an architectural

role in designing the data model, the database technology supports
the effective implementation of this model. The presented
architecture structures information along two dimensions: vertically
with the partitioning of the set of clinical components as hierarchically
or ganized horizzontally with the introduction of a blackboard
structure for distribuited information pro cessing. Accordingly, the
clinical information system takes the form of a distribuited database
system, whose purpose is to automate clinical tasks and to support
administrative and financial functions.

Database and Expert Systems Applications.pdf

UnPatient has been implemented in Prolog-DB, which allows to

514.
combine all pro's oflogic programming and processing of knowl edge
stored in data base management systems [8]. Prolog-DB is based on
a relational data base management system (db++) and Prolog. The
entire object hierarchy has been mapped to db++ relations.

Database and Expert Systems Applications.pdf

515. Database management systems can model complex real world

situations, but do not have "the single coherent inter face to the
database which is the hallmark of hypertext"[l]. In this paper, we
propose a general hypertext-based documentation support system,
InterSect, which uses an object oriented data base as an information
repository to support both the definition and manipulation of complex
document structures. This proto type provides both hypertext and
database views of documents, together with a versioning mechanism
driven by a user defined schema for the overall document structure.

Database and Expert Systems Applications.pdf

516. Special Requests It seems advisable to stress here that the requests
to such an index system are quite different from the requests to (e. g.
prosopographical) historical data bases [10]. On the one hand, a
printed index has to satisfy fewer kinds of requests than a historical
database normally has to, on the other hand there are many other
 special requests: There are very particular demands in sorting, e. g.
in sorting certain ambiguous strings in a manner which is different
from the normal sorting rule applied to a special task. It can be that
many different sorting rules must cooperate in a complex system: e.
g. alphabetical, chronological and the classifying of grammatical
types of word connections.

Database and Expert Systems Applications.pdf

On the other extreme, the knowledge engineer has a deep knowl

517.
edge of the data model, the database context and of the data manip-

Database and Expert Systems Applications.pdf

518. Fig. 2: System Architecture of the First Prototype The database

contains economical and technical data of working panels and old
workings. These data originate from the database ZPD1 of the
German mining company Ruhrkoh Ie AG, which is supporting the
ongoing research work.

Database and Expert Systems Applications.pdf

Data re sources, servers, and applications are heterogeneous. The

519.
environment needs to support frequent changes and addi tions to
these data and computation resources. Information and control often
has to be exchanged among different infor mation bases.

Database and Expert Systems Applications.pdf

Creat- ing the perfect plan for learning ontology from RDB requires a
520.
clear understanding of the domain area, the problem to be solved,
and scoping of the data sources to be used. Answering these
questions clarify the problem definition and help us to select the
appro- priate database that can be used in later phases. The Table 5
exhibits a list of compe- tency questions ( CQs) that represent
informal questions that the ontology must be able to answer [41].

A novel approach for learning ontology from Relational db.pdf

 The underlying model for building application data is a domain
521.
specific ontology. According to this ontology application data can be
semantically interpreted and enables retrievals for context-based
information, e. g. by using synonyms.

Lecture Notes in Artificial Intelligence.pdf

Many standardiza- tion efforts aim at solving these problems [2].

522.
Standards play an important role for ensuring a common
understanding of transferred data among heterogeneous application
systems [3]. To achieve effective communication, not only technical
interfaces are required, but also common semantics for exchanged
data.

Lecture Notes in Artificial Intelligence.pdf

More, an ObjectProperty will be created and its domain will be the

523.
event class and as a range the subclass of the event class. The
capability to translate RDB trigger in an ontology form is an optimal
feature for an ontology because domain knowledge reflects the
world, as it is constantly evolving. Although the fact that the data
inside the database are not translated is a significant shortcoming.

Ontology learning from relational database a review.pdf

(2013) propose a more detailed translation of the constraints that

524.
may be present in a database, but they completely leave out the
possible translation of hierarchical structures, which are fundamental
in the construction of complex, well-structured ontologies. Moreover,
as can be seen from Table 1, such works, together with the Hazber et
al. (2016), give in output an OWL-Full Ontology (because of the
inverse functional data property translation of the unique constraint),
which, since cannot be used with a reasoner, may be less useful than
ontologies written in OWL-DL.

Ontology learning from relational database a review.pdf

Overcoming semantic heterogeneity is typically achieved in two

525.
steps, namely: (i) matching entities to determine an alignment, i. e., a
set of correspon- dences, and (ii) interpreting an alignment according
 to application needs, such as data translation or query answering.
We focus only on the matching step.

Ontology matching state of the art and futur.pdf

526. However, these constraints or automatisms are not generated by

these methods. In this case, the RDBMS cannot guarantee intrinsically
the integrity of the data as described by the ontology. Yet, the
integrity of the database can be evaluated externally if the
information is present in the metadata, and internally if explicit
constraints were generated.

Generating_Relational_Database.pdf

527. Regarding structure reversibility and tuples reversibility, interesting

work has been done in the reversibility aspect. That means that the
methods are used in a more global context such as ontology data
access, data acquisition, and data extraction.

Generating_Relational_Database.pdf

528. Table 1. Correspondence of data types between database and OWL

ontology

An Improved Ontology Learning Algorithm from Relational Schema.pdf

In this paper, we discussed these fundamental problems and outlined

529.
the architecture of our MDBMS system. One contribution of our paper
is the formulation of a partial matching algorithm that uses domain
knowledge, represent ed using an object-oriented data model, and a
weight rank ing system to assign weights to different multimedia data
stored in a database and selects those multimedia data that partially
match a given user query description. Our parser. unlike others,
provides an interpretation of natural language descriptions needed to
achieve an intelligent retrieval of multimedia data.

Database and Expert Systems Applications.pdf

For interactive queries, the system should allow vague queries and
530.
query for mulations that are independent of the specific structure of
 the data and its representation. For vague queries and imprecise
data, methods developed in information retrieval can be ap plied.
Heterogeneous data structures can be handled with con cepts from
object-oriented database management systems.

Database and Expert Systems Applications.pdf

The aspect of reusing visual objects for different target applications

531.
and probably different VUIs brings up the topic of object
representation. In the case of visual objects there are not only logical
properties such as descriptors and references but also physical
attributes such as dimensions (image x-size, y size), positions
(textual annotation at <xt, yt», or pure format data (image I is of type
ROB, - TIFF). These physical properties must be considered for the
final object instantiation in a specific target VVI, but for all the
intermediate steps of object and database definitions, the logical
characteristics are the decisive ones.

Database and Expert Systems Applications.pdf

532. 4 Information Requirements The information system H.I.T.t, which

is used in the after-care centre of the Stadtische Kliniken Oldenburg
and other clinics in Niedersachsen, does not meet modern
information-technical standards in decisive features (operating
system, data model, data manipulation language). Despite the
existence of an ANSI standard, the underlying programming
language MUMPS2 ensures no portability with regard to various
target computers and limits the flexibility with regard to desirable
possibilities of analysis, e. g. for epidemiologic cancer research.

Database and Expert Systems Applications.pdf

533. 569 6 Use of the INEKS Database for Epidemiological Cancer

Research There are analyses which do, for the most part, obtain their
data from an underlying database, but would like to use in addition
DB-external data, for example from files or main memory resident
data structures such as records, arrays, or which shall be entered by
keyboard for test purposes. This DB-external data is often not
integrated in the DB because it is, for instance, only of temporary
interest (and would not justify a laborious re organisation of the DB
schema), or because it would expand only a part of the DB by
 additional attributes and raise consistency problems (zero values!)
for the remaining DB section.

Database and Expert Systems Applications.pdf

534. Gupta, Condit, and Qian [41] introduce the BioDB multi-model
system, which is able to manage heterogeneous biological data and
information where different query processing operations are used for
different categories of biological data. The system operates with
relational, graph-based and tree-based information models, which
are extended by ontological annotations providing meaning for these
data. As the authors claim: “An ontology-enhanced system is a
system where ad hoc data is imported into the system by a user,
annotated by the user to connect the data to an ontology or other
data sources, and then all data connected through the ontology can
be queried in a federated manner.”

Formal Ontologies in Information Systems.pdf

The information in existing management system of enterprises is

535.
very critical to management. These management systems can
independently complete some work, like service management,
decision making, and knowledge extracting for intelligent reasoning.
Traditional information resources are mostly based on relational
database, which stored most of enterprise data and contains
numerous history data ( Küpers, 2015; Lin etc.,

An Improved Ontology Learning Algorithm from Relational Schema.pdf

536. Table 3: form table ATTRIBUTE meaning I form 10 identifier of form I

item pointer pointer to top item(field,comment,button) I . ~. : 5.3 Field
elements Field elements show field types, length and nominated data
to com pose forms. ( Refer to Figure 6 , 7.)

Database and Expert Systems Applications.pdf

Despite a popular belief in object-oriented data models, we found

537.
that the formats of most VLSI/ CAD data are relational in essence.

Database and Expert Systems Applications.pdf

 The meaning of data cannot be derived from the format of a fIle
538.
alone, since meaning mostly is distributed between data files and the
corresponding program. E. g. : a record of a file must not be
considered as an isolated object, if the ordering in the file is crucial
for the ordering in which the records must be processed by a
program. Therefore the formulation of integrity constraints and
consistency checks have been an important part within the
integration of several files concerning the same objects of the
application domain.

Database and Expert Systems Applications.pdf

The MCC database program [15] also undertook several multimedia

539.
projects and identified requirements for a data model suitable for
multimedia ap plications. [10] has developed a framework to classify
and compare the different multimedia projects.

Database and Expert Systems Applications.pdf

4.1 Natural Language Descriptions As mentioned, we propose to

540.
perform retrieval of multi media data by matching the natural
language descriptions with query specifications. We believe that
unrestricted nat ural language processing is very difficult to achieve
given the AI technology today. We found that the language needed
to describe multimedia data is much more formal than ev eryday
English.

Database and Expert Systems Applications.pdf

541. In addition, many of these prefixes are redundant. The greater the
number of images is, the larger the redundancy is. Because, we are
in presence of a very large database, this kind of data distribution
and use of the quad tree generates an important run time and
processing complexity.

Database and Expert Systems Applications.pdf

For example, as shown in Fig. 1, the young fuzzy-object has the

542.
following: objectName: young contents: {(0.7, a pointer to 30), (0.9, a
 pointer to 25), ... } This data structure enables to handle the fuzzy
object as a single object.

Database and Expert Systems Applications.pdf

543. In the discussion below, we assume that the export schemas are
relational. Hierarchical, network, and other data models used by the
LDBMSs can be accommodated by including neces"ary dat. a model
t.

Database and Expert Systems Applications.pdf

A complete multiversion ( MV) schedule over a set T of trans actions

544.
is a triple where: T p( T) = U:~ ii’ Ti is a set of all database
operations involved in the transactions of the set Tj <mv. p2 U:~ ii’
<i is a partial order on Tp(T) compatible with the partial orders ~
databa~ operations specified by the trans~tionsj Rdmvsp : Tp(T) ......
Tp(T) is a reads-from relation over Tp(T), which de termines the
versions of data items read by transactions of T.

Database and Expert Systems Applications.pdf

7.3 Maintaining Temporal Information Several authors discuss the

545.
problem of infinite data arising in the con text of temporal query
processing [1, 14]. This problem also occurs in the context of the
query language of this paper.

Database and Expert Systems Applications.pdf

4.2.2 Multilinguality Apart from all the basic conceptual problems

546.
modelling, it is also necessary to decide on a language used within
the modelling effort. For the generalized data model of the
organization, it is obvious that it has to be described in the "official"
common language of the organization. For the lower-level models,
the decision clearly depends on organizational standards.

Database and Expert Systems Applications.pdf

When an order arrives, the possible production sites are polled to

547.
determine their capacities, costs, schedules, etc. Given this data and
 the customer's expectations, distributed production schedules for the
order are produced for the vari ous sites. We have implemented
database support for execut ing such plans; we have not attacked
the scheduling problem itself.

Database and Expert Systems Applications.pdf

We have identified the following three classes of mediators ( Figure

548.
1): Task Models We break out domain knowledge now hidden in
applica tion programs and store them in a special kind of mediator
which implements sharable and inspect able domain knowl edge
bases. We call such a domain knowledge base a Task Model. The
task models allow us to maintain knowledge more easily by storing it
in these limited-sized and special- ized knowledge bases.

Database and Expert Systems Applications.pdf

549. Also,it is adaptable to changes or advances in computer technology.

ln particular, the standardized opera ting system environment and the
modular system structure allows the integration with existing clini cal
subsystem or the integration with new subsys tems considering
advances in computer technology (i. e. workstations, multi-media
information repor ting, picture archiving.,

Database and Expert Systems Applications.pdf

550. According to [1], which describes the successful use of an

experience repository in a small software company, detailed data on
its use and structure can be used to better understand how
experience supports activities in the company. This can in turn lead
to improvements in experience management concepts, techniques
and tools.

Lecture Notes in Artificial Intelligence.pdf

551. Fig. 3. Domain model generates various possibilities of scenarisation

4.2 Defining the Scenarios from the Main Activities Model The
general scenario can be further detailed as follows: − The patrimony
model allows defining the different scenarios elements such as global
prerequisites or the global teaching objectives the main activities
model and the different activity steps (cf. Fig. 3). The organization of
 steps appearing in the general structure of learning activity is the
same that the one describes in the main activity model because the
great phases defined in the main activity model are articulated
between them by exchanges of data, of documents and others of
where a "functional" organization that answers to the convenient way
of the activity and, exactly, that must be presented as such to learner
who wishing to learn the profession.

Lecture Notes in Artificial Intelligence.pdf

Methods that create an OWL-Ontology with concepts and properties

552.
and that allow the user to access the data present in the database,
for example converting SPARQL queries into SQL queries, but which
do not directly import instances into the ontology.

Ontology learning from relational database a review.pdf

553. 3 APPLICATIONS Ontology matching is an important operation in

tra- ditional applications, e. g., ontology evolution [23], on- tology
integration [24], data integration [25], and data warehouses [26].
These applications are characterized by heterogeneous models, e.

Ontology matching state of the art and futur.pdf

11 USER INVOLVEMENT In traditional applications, the result of

554.
matching performed at design time is screened by human users
before being accepted. However, the overwhelming size of data may
render this task difficult. In dy- namic applications, users are
generally not ontology matching specialists who can be asked to
inspect the alignments.

Ontology matching state of the art and futur.pdf

555. A. General Observation The OWL language is the most popular

ontology language. However, the methods differ from one another
according to the ontology constructs taken into account in the
conversion process. The common ontology constructs used are:
classes, objects properties, data properties, subclass axioms with
simple restriction (some, exactly, min and max), functional properties
characteristic as well as the domain and the range of properties.
 Generating_Relational_Database.pdf

Advancing Clinico-genomic Trials on Cancer ( ACGT) is an EU co-

556.
funded project focusing on improving medical knowledge discovery
with the assistance of grid technologies and ontological engineering.
Brochhausen et al. [44] present ACGT-MasterOntology ( ACGT-MO)
—an application ontology which supports data integration across
different terminologies used by different countries, disciplines, and
languages.

Formal Ontologies in Information Systems.pdf

4) The existing relational data model did not take into consideration
557.
of the associated conceptual model of one to one.

An Improved Ontology Learning Algorithm from Relational Schema.pdf

SNAP and Pasta-3 provide the user with flexibility both in schema
558.
definition and manipulation. Second, since many research prototypes
are based on either entity relationship, or semantic, or object-
oriented data models, they should base the data manipulation on the
same schema representation as the one used at schema definition
time. as suggested in [5, 13]. However. only some prototypes
provide data manipulation facilities close to those defined for
relational databases.

Database and Expert Systems Applications.pdf

559. Liu and Gao (2018), proposes a method for learning ontology from
an RDB called WN_ Graph. Since this method is based on Dadjoo and
Kheirkhah (2015) and the authors do not specify anything about the
input and output of their model, we can deduce that, like ( Dadjoo
and Kheirkhah 2015), the input of the system is a relational database
writ- ten in SQL Data Definition and that its output is an ontology
model in the OWL structure in which only the concepts are imported
but without any instances resulting from a map- ping of the database
data. To provide a better hierarchical relationship between concepts,
the proposed method uses the intermediate conceptual graph model
 combined with WordNet so that the approach can extract more rich
seman- tic relationships from the RDB.

Ontology learning from relational database a review.pdf

560. The progress in database theory research - especially object oriented

approaches - allows for the first time to profit from database
technology in such a complex domain. A first prototypic
implementation including essential mining objects and several
applications (topological scanning) was tested with existing
productional data from one mine. It yielded promising results with
respect to perfonnance and fruitful experiences for the future overall
system development.

Database and Expert Systems Applications.pdf

While there are numerous, well known approaches and metho

561.
dologies for traditional, relational database design and informati on
systems modelling (e. g., the entity relationship approach [9], se
mantic data models such as RM/T [11] and SDM [14], dependency
theory [16]) no such generally accepted methods exist for building
knowledge-based information systems. While the methods mentio
ned above have made it easier for a user to model an information
system, several concerns still remain unresolved.

Database and Expert Systems Applications.pdf

88 By making available for each user / each application all the

562.
relevant data of the mine, it is guaranteed that they always access
the most actual state. There should be no time lag between data
input by one department and data access by another one. This fact
improves data reliability, too.

Database and Expert Systems Applications.pdf

563. 3 Knowledge Component The primary purpose of the knowledge

component of the system is to determine the exact location a fault
occurred once it has been detected. The diagnosis process takes as
input a fault, the location at which the fault was detected (by a
sensor), and a fault type (determined by the Data Analysis Module).
Based on this input, the current state of the device which is
 determined by the sensor data, and knowledge of the structure of the
device, the diagnosis process must determine which component, or
compo nents, could have failed.

Database and Expert Systems Applications.pdf

The two main tasks performed by the DMM are the insertion of the
564.
sensor data into the database, which must be performed in real-time,
and the handling of que ries from the KEM during the diagnosis of a
fault. Our

Database and Expert Systems Applications.pdf

3 General Architecture Although DB offers a unified view of

565.
meteorological data, it comprises different modules, each having their
own data representation scheme. The current implementation of the
database is composed of four modules. There is one module for each
type of data (soundings, surface based observations, and gridded
data) and one controlling module which inter faces with the client
(see figure 1).

Database and Expert Systems Applications.pdf

566. Now, consider the structure of MVSSG(mvs(r)) from the point of view
of the relationship between an active transaction Tk (represented by
the transaction node Tk) and a data item x represented by a
sequence of versions nodes X~" X~2’ ,X~g’ From the point of view of
Tk, versions of x can be divided into three classes: those which are
predecessors of Tk in the graph, those which are successors of Tk in
the graph and those which are unrelated to Tk.

Database and Expert Systems Applications.pdf

567. This paper proposes a query language and data organisation for
large hyperbase systems. The language allows queries to involve the
links as well as the text. The architecture allows the data to be
efficiently indexed, as well as supporting multiple users.

Database and Expert Systems Applications.pdf

 No type definitions for node and link A node in the hypertext systems
568.
is a storage unit for storing a collection of data items. We cannot find
out detailed information about its structure. Every node is treated as
the same type.

Database and Expert Systems Applications.pdf

It is advantageous if the active DBMS can do this kind of dynamic

569.
filtering before notifying the application, in order to decrease the
frequency of notification for intensively up dated data. The application
can dynamically change the dif ference interval to decrease the
reactivity if it is performing some change-intensive task. Dynamic
filtering is required, for example, by real-time monitoring AI systems
where the tracker initiates time consuming reasoning activities [21].

Database and Expert Systems Applications.pdf

570. THE MEDICAL REPORT While the blackboard as whole provides a

structure for decentralizing the processing of a common set of
relevant clinical data, its primary objective is to ensure that each
clinical department can identify which particular report must be pro
cessed. This is achieved by associating each patient with a code
called nosological number, which acts as a key to collect all patient
related informa tion. This means that there is a unique name and
code associated with each patient item on each black board.

Database and Expert Systems Applications.pdf

El Alami A, Bahaj M. The migration of a conceptual object model

571.
COM (conceptual data model CDM, unified modeling language UML
class diagram...) to the Object Relational Database ORDB. MAGNT
Research Report ( ISSN. 1444–8939). 2:318–32.

A novel approach for learning ontology from Relational db.pdf

572. VIKAMINE offers different search options for automatic subgroup

discovery and various interactive visualizations for active user
involvement. Both tools operate on the same data base. When the
user discovers something unexpected/interesting in the data using
 the simple tool, then, after switching to VIKAMINE, these unexpected
features can be inspected and analyzed in detail, e.

Lecture Notes in Artificial Intelligence.pdf

573. PHRs offer the possibility of medical records that can be easily
accessed and annotated both within and between organizations3, 4.
Post-genomic clinical studies and the application of data mining
methods imply stan- dardized patient data and standardized
terminology being accessible to knowledge discovery tools5. Early
attempts on standardizing medical vocabulary show the com- plexity
of this task6-8.

Lecture Notes in Artificial Intelligence.pdf

Integrated Clinical Pathways in Health Information Systems 357 tion

574.
information. Execution time of actions is determined by transition
conditions. Activities can relate to resources and can be coupled with
computer applications or functions by a set of function parameters
and by sharing system data.

Lecture Notes in Artificial Intelligence.pdf

575. 4 Further Knowledge Projects In the meantime, further knowledge

projects were initiated, for which a demand in ISB AG exist. For
example applications for the standardized maintenance of em-
ployee’s profiles and project warrant of apprehension were conceived
and already partially integrated in PlanIT. On this occasion, the user
is supported with the input and maintenance of the individual data by
standardized and centralized entries.

Lecture Notes in Artificial Intelligence.pdf

576. On the other hand, the majority of current Web content is dynamic
content powered by relational databases ( RDBs) [4,5], in which
abundant domain semantics has been (implicitly) encoded.
Therefore, it is valuable to investigate approaches and tools for
extracting such domain semantics from RDBs and then constructing
OWL ontologies in an automatic or semi- automatic way.
 OWL Ontology Extraction from Relational.pdf

[12] L. Lubyte and S. Tessaris, “ Automatic extraction of ontologies

577.
wrapping relational data sources,” in Proc. of the 20th Int’l Conf. on
Database and Expert Systems Applications, LNCS, vol. 5690, S. S.
Bhowmick, J. Küng, R. Wagner, Eds., Heidelberg: Springer Berlin
Heidelberg, pp. 128–142, 2009.

OWL Ontology Extraction from Relational.pdf

578. 791 Sonika Malik et al. / Procedia Computer Science 70 ( 2015 ) 785
– 792 which means updating the knowledge base through mapping
from all the relevant ontology or distributed data available online.

Super Ontology.pdf

579. else { ( ) ooEntityTab T true← ; continue}; if T has exactly two foreign

keys 1 2( ), ( )fk T fk T and one primary key ( )pk T such that 1 2( ) ( ),
( ) ( ) ,fk T pk T fk T pk T= ∩ = ∅ 1 2( ( ), ( )), ( ( ), ( )), Ref fk T pk R
Ref fk T pk S R S≠ then if many values in ( ( )) Data pk T relate to
one value in 2( ( )) Data fk T then { ( ) omRelTab T true← ; continue}

OWL Ontology Extraction from Relational.pdf

Abstract data types provides modularity, extensibility and reusability.

580.
Views provide logical independence and allow to have different views
on the same objects. Views also give an intuitive definition of
applications in terms of inter-related tasks; (3) Serendip is safe.

Database and Expert Systems Applications.pdf

581. [12] Katz, R.H. A database approach for managing VLSI design
data. Proc. 19th ACM/ IEEE Design Automation Conference, 1982,
pp. 274-282.

Database and Expert Systems Applications.pdf

582. Based on a flow-chart description of the planning process it supports

a systematic construction of mine structures, avoiding faults and
 omissions, and identifIes critical points in the design. The system is
implemented with Nexpert Object, having a SQL interface to
databases, but no interface to FORTRAN. Since most of the programs
in the domain mine planning are implemented in FORTRAN 77 and
the data are structured accordingly a database is necessary
providing access to the existing geometric data via a SQL interface.

Database and Expert Systems Applications.pdf

We can abstract the contents of multimedia data into words or text

583.
and use the text description equiva lent of the original multimedia
data to match a user query. This principle we will use in designing a
MDBMS to handle multimedia data for different applications.

Database and Expert Systems Applications.pdf

ically switches to approximate match. As the solution to the natural

584.
language part of a query, the query processor re ceives links to the
qualified multimedia data. After combing them with the results of the
conventional subquery the final results arc retrieved by the Data
Access Subsystem.

Database and Expert Systems Applications.pdf

When a query is received from the user, the parser sepa rates the
585.
natural language specification into smaller compo nent groups,
namely subject noun, verb and object noun phrases. Each of these
will actually become predicates. When these predicates match
exactly with the predicates in the descriptions of certain multimedia
data, those multime dia data will be retrieved.

Database and Expert Systems Applications.pdf

 Most queries involve both media and formatted data. For the media
586.
part of the query we use our intelligent matching algorithm which is
directly processing natural language captions. For conditions on
formatted data, natural lan guage expressions are mostly too
imprecise to be directly processed.

Database and Expert Systems Applications.pdf

587. An illustrative example The SCHEDULER is an excellent basis for the

implementa tion of data parallel deduction. We now illustrate the
basic ideas by a description of the computation of the transitive
closure of a graph G. The following two rules define our pro blem:
TC(X,Y) <- G(X,Y); TC(X,Y) <- TC(X,Z), G(Z,Y); We assume, that the
graph G is stored in a database rela tion G( Gl,G2), and we want to
compute its transitive clo sure and insert in into relation TC. Note that
Horn logic "predicates" and database "relations" are synonyms in our
context.

Database and Expert Systems Applications.pdf

588. Also, we will describe our prototype fuzzy OODB system, which was
implemented on the commercial OODBMS Versant[8] and is
currently running. An exam ple database is a collection of textual
data and PostScript image data concerned with movie information.
By using this example database, we will briefly show the function ali
ties of our prototype system.

Database and Expert Systems Applications.pdf

589. Constraints can be enforced by only examining data that has actually
been changed by a transaction; this technique requires the
manipulation of differential sets [ Si87]. Integrity con straints can be
rewritten such, that they can be evaluated more efficiently. In the
case of distributed database systems, the in tegrity constraint rules
can be distributed over the fragments of relations, thereby avoiding
the overhead of reconstructing global relations.

Database and Expert Systems Applications.pdf

 4 Problem formulation Any multiversion concurrency control
590.
algorithm, besides decid ing at each moment whether to grant, delay,
or abort an arriving read or write operation of a transaction, and
which of the ex isting versions of a data item accessed should be
presented to a read operation, must also decide which of the existing
versions of a data item, if any, to overwrite. This latter decision is
made by a version control manager, which realizes a given overwrite
protocol. Most of version control managers designed so far or
proposed for use in MVDBSs apply the same overwrite protocol
which we called natural overwrite protocol [2,3,5,7,13].

Database and Expert Systems Applications.pdf

Dynamic overwrite policy (dop) 1. If the total number of valid versions

591.
of all data items is less than the system-imposed upper bound K,
then a ver sion created by a committed transaction is added to the
database, otherwise, 2. it overwrites a version x;. E DB according to
the following rules: if the set nTkEM U NVt( Tk) is non-empty, then x;.
is the oldest version contained in the set if the set nTkEM U NVt( Tk)
is empty, then x;. is cho sen from one of sets U NVt( Tk), k = 1,2, ...

Database and Expert Systems Applications.pdf

592. For data sorts the semantics is fixed once for the complete runtime of
the database whereas the interpretation of the object sorts varies in
time. For example the square of the number 5 is always 25 whereas
ti-J salary of the P E RSO N ’ Smith’ may change.

Database and Expert Systems Applications.pdf

1 Introduction As information and communication functions grow,

593.
more elabo rate human-computer interfaces need to be made
available to fmal users. For some time now, progress has been made
towards of fering easy access interfaces to a data base for non-
specialist users. Interfaces based on interactive graphical
manipulations constitute an important class because of the obvious
user-friend liness they can offer.

Database and Expert Systems Applications.pdf

 Each ST is also univocally associated with a data type that
594.
expresses the ag gregative function that has generated the table
summary values.

Database and Expert Systems Applications.pdf

420 represent the modalities of the variable "regions". "N°" identifies

595.
the data type of the table and it is used to indicate that the values are
absolute values obtained by count operations.

Database and Expert Systems Applications.pdf

In particular we are demonstrating active mediatoTJ, where the

596.
application instructs a mediator to actively moni tor databases for
change in information that the application depends on. We identify
how mediators can support appli cations that are sensitive to change.

Database and Expert Systems Applications.pdf

Transmission of H.I.T. data The data stored in the information system

597.
H.I.T. should be transferred into the INEKS database. Therefore it
was necessary to develop a program because the internal data
representation of MUMPS was unknown. The H.LT. data are given out
and stored in a file named transfer file created by an MUMPS program.

Database and Expert Systems Applications.pdf

SC was developed continuously on the basis of user feedback. It

598.
contains about 430 questions, 140 symptom abstractions and 230
diagnoses. The analysis of the data base of 770 consecutive cases
exhibited a mean of 61 answered questions per case with an
average of 20 symptom abstractions and 6 diagnoses inferred by the
program.

Lecture Notes in Artificial Intelligence.pdf

Table 1. Dataset characteristics Name Type Patient and

599.
hospitalization related sex { Male, Female} age [0;72], mean 29.8
recurring stay { True, False} days of stay in NSI before test [0;317],
 mean 87.5 days of stay in ICU [0;237], mean 34 days of stay in NSI
before specimen was received [0;169] mean 31.6 bacterium is
isolated when patient is in ICU { True, False} main department
{0,…,9} department of stay {0,…,11} Pathogen and pathogen groups
pathogen name { Pat_name1, …, Pat_name17} group1 { True, False}
… … group15 { True, False} Antibiotic and antibiotic groups antibiotic
name { Ant_name1, …, Ant_name39} group1 { True, False} … …
group15 { True, False} sensitivity { Sensitive, Intermediate, Resistant}
Each bacterium in a sensitivity test in the database is isolated from a
single specimen. that may be blood, liquor, urine, etc.

Lecture Notes in Artificial Intelligence.pdf

Besides, in this paper we demonstrated how complex many-sided

600.
analysis of a real-world problem could have been performed, how
application of different DM techniques has helped us to understand
better the nature of the data that represents the problem.

Lecture Notes in Artificial Intelligence.pdf

601. The sending of data is successfulness. Actors BrenchOperator,

HeadQuarterSystem. Use Case Extends Nn Use Case Uses Check
Correctness/ Completeness Use Case Inputs Name, address, offered
books list (in case the user is a vendor) with specifications: title,
author, publisher, language, publishing year, ISBN.

Lecture Notes in Artificial Intelligence.pdf

Semantic Interoperability Among Models 555 based on model A in

602.
order to make it usable for systems that expect data to be based on
model B. The following figure shows the mapping and its application.

Lecture Notes in Artificial Intelligence.pdf

Basically, further research is needed in at least two areas. Firstly, the

603.
specifics of behav- ioral aspects have to be clearly identified in order
to either adapt the schema integration process for static aspects as
given e. g. in [11], or to come up with an integration process for
behavioral aspects building on a specific common data model.
 Lecture Notes in Artificial Intelligence.pdf

CONCLUSION We presented an approach to deductive query

604.
optimization which allows for abstract data types and arbitrary user-
defined functions within the frameworlc of an extended relational
model. In this way we support some of the features of object-oriented
systems, in addi tion to deductive inferencing, Functions may appear
in both rules and queries, and the query optimization process uses a
combination of tenn rewriting and distribution of selections. This
provides a clean integration of functions into both the language and
the optimization algorithm.

Database and Expert Systems Applications.pdf

e. the translation from relational symbols to OWL class and data range
605.
identifiers, and all of the operations for creating datatype/object
property identifiers in Step 2 of algorithm SchemaTrans can be
simultaneously made as sub-operations when creating the OWL class
and property axioms in Step 2, we can ignore these sub- operations
and consider only the creation of class and property axioms in Step 2
to be the basic operations of algorithm SchemaTrans. Therefore, the
basic operations of this algorithm are counted as follows: TABLE II.
A LIST OF AUXILIARY PREDICATES Category Symbol Meaning
Entity tables ( ) normEntityTab T T is a normal entity table ( )
weakEntityTab T T is a weak entity table ( ) ooEntityTab T T is an
entity table containing a one-to-one binary relationship ( )
omEntityTab T T is an entity table containing a one-to-many binary
relationship ( ) subEntityTab T T is a subtype entity table ( )
superEntityTab T T is a supertype entity table Relationship tables ( )
naryRelTab T T is an n-ary relationship table.

OWL Ontology Extraction from Relational.pdf

606. The purpose of this paper is to present a methodology for rep

resenting and manipulating polymorphic object state and behav- ior
in the process of modeling data intensive applications. In this paper
we view the object-oriented paradigm under a modeling driven
perspective where the use of objects amounts to adopting data
modeling constructs that are direct mappings of problem domain
 concepts. Evolution of objects is supported by allow ing them to
undertake or relinquish numerous ephemeral roles.

Database and Expert Systems Applications.pdf

607. The ES is used for deductive filtering of the data to be stored in the
database and for both the user and sub queries. The inflexibility
arises because the ES is written to interact with the DB rather than to
implement the domain knowledge of an expert.

Database and Expert Systems Applications.pdf

The resultant data structure, which is similar to the multilist

608.
representation of links in a network model [14], is shown in Fig. 3-2.

Database and Expert Systems Applications.pdf

This paper makes three contributions. The first contribu tion is that
609.
context description of multimedia data is possi ble using natural
language captions which can be interpret ed automatically using
domain dependent knowledge. The second contribution is the
formulation of a general scheme to retrieve multimedia data with
special emphasis on ap proximate match.

Database and Expert Systems Applications.pdf

In the domain of IR there had been early interest in using AI

610.
techniques [14]. More representative bf modem attempts are the
IRUS system [2] which is designed for processing heterogeneous
data bases through natural language queries and the IOTA system
[4] which tries to improve the quali tative performance of IR systems
in replacing keywords by noun groups. The approach we propose is
somewhat differ ent from the intelligent IR systems proposed so far.

Database and Expert Systems Applications.pdf

611. The con- Raw Data ( Matrix of Pixels in Raster/ Bitmap FonnaL)
Description Data (abstracted content of image using text) Figure 1 :
Example for the Multimedia Data Format Data Access Subsystem
Intelligent Retrieval Subsystem Figure 2: Architecture of MDBMS
 System tents of a multimedia data is described by the description
data. Description data cannot be automatically derived by the
computer given the technology today. We assume that users will
supply the description data for multimedia data in a natural language
form.

Database and Expert Systems Applications.pdf

612. 4.3 Natural Language Interpretation The parser translates the· text
description into a set of predicates called meaning list, thereby
reducing impreci sion and ambiguity of the natural language
descriptions considerably. These predicates state facts about the real
world entities involved with multimedia data like their properties and
relationships. As in most parsing methods, we chose the use of first-
order predicate calculus. as a formal representation of the description
data.

Database and Expert Systems Applications.pdf

Fur thermore, they are not able to formulate requests by using a

613.
formal query language. Hence. they need an interface that presents
the in formation according to their real-world vision, and that allows to
access and manipulate data by means of easy-to-use commands.

Database and Expert Systems Applications.pdf

614. An efficient and flexible, fully compiled rule language is included as

well as an advanced graphics library. TOROS knows persistent
objects and provides operations for storing and loading object data.
TOROS can run under Unix, and a DOS version is under development.

Database and Expert Systems Applications.pdf

615. In traditional Hypertext systems, in contrast with conventional in

formation systems, there is almost no conceptual description of the
stored data. The weaknesses of such an approach have been discus
sed by several authors ([7][15]). There seems to be a growing need
to be able to support a conceptual description in relation with
Hypertext or documents, especially the combination of structured
documents with Hypertext applications looks promising.
 Database and Expert Systems Applications.pdf

DB integrates the three modules and offers a uniform view of the

616.
different data sources. This in turn allows highly flexible data access.
DB can be accessed either interactively by the user, or within an
application written in "0" or FOR TRAN.

Database and Expert Systems Applications.pdf

312 schema integration. A traditional approach to query

617.
decomposition and optimization, based on estimation of volume of
data transfer and cOlnnluuication costs, is suitable for systenlS in
which queries are formulated assuming a global integrated schema.
Depending on the scope of functions provided by the sit.

Database and Expert Systems Applications.pdf

In stead of specifying a global conceptual scheme which inte grates

618.
all the component schemes, an incremental approach is used for the
stepwise integration of partial knowledge. The user is supported by
facilities to express identified relations between data elements,
particularly those between data el ements of different databases.
Formulating those relations as extensions of the semantic level leads
t? a partial global conceptual scheme that is relevant to the sriecific
application domain, subsequently called " Semantic Background
Model".

Database and Expert Systems Applications.pdf

For example, having access to the raw observatory data about an ob

619.
ject is as important as the related scientific reports. Thus, a DBPL for
this environment should be prepared to accumulate a large type
catalogue and be an open language to access remote resources.

Database and Expert Systems Applications.pdf

There are several suggestions to extend SQL by transitive closure

620.
functions which support- nested objects [~. This article presents
another app roach: - The implementation of the DOAG structure as
 rela tional tables is kept. - The relations are transformed to
appropriate dynamic data structures.

Database and Expert Systems Applications.pdf

621. Thill concept has to be linked to all the other entity types which make
up for the description of any individual operating unit. Thus, the
generalized data structure is able to capture structurally different unit
data structures within just one model. Depending on the particular
organizational unit, the linkage to certain other concepts may not be
filled.

Database and Expert Systems Applications.pdf

622. A large amount of data was collected, which is currently being

analysed as part of a subsequent investigation into the application of
quantitative model building. Currently a commercial tool, Decision
Explorer™ by Banxia™ [20], is being assessed for use in this case
study.

Lecture Notes in Artificial Intelligence.pdf

Data storage in a data warehouse: the basic idea of a data

623.
warehouse is to store the relevant data for decision makers in a
dedicated, homogeneous database. An important characteristic of
the data warehouse is the integration of heterogeneous, distributed,
internal and external data. This covers the physical storage of data in
a single, centralized data pool, and it also covers the subject-oriented
clustering of data organized by business processes, such as sales,
production, or finance.

Lecture Notes in Artificial Intelligence.pdf

624. In addition, when a primary key must be translated an inverse

functional data property is created: this means that the gen- erated
ontology will be in OWL-Full Language. Unfortu- nately, as written
before in the [24] comment, OWL-Full typically is not fully complained
with reasoning activity, and for this reason, an ontology modeled in
OWL-Full may be less useful than an OWL-DL Ontology in terms of
 possible inference on the data. The authors evaluate their prototype
application on a case study based on a small dataset.

Ontology learning from relational database a review.pdf

625. The adjacency list is easy for a person to understand and write
programs for processing it, but its effective implementation is
possible only in high-level programming languages (C#, Java,
Python) with the built-in support for dynamic data types, which
negatively affects the performance of such programs in comparison,
for example, with the C++ programming language.

Graph to RDBMS.pdf

626. Part of the task structure for di- agnosis is shown in Figure 2. The
diagnosis task can be viewed as an abductive task, the construction
of a best explanat ion (one or more dis- orders) to explain a set of
data (manifestations). The task structure shows three typical
subclasses of abductive methods: Bayesian, ab- ductive assembly
[19] and parsimo- nious covering [30].

Task-Structure Analysis for KM.pdf

627. Mittal, S. and Chandrasekaran, B. Patrec: A knowledge-directed

data- base for a diagnostic expert system. Computer 17, 9 (1984),
51-58.

Task-Structure Analysis for KM.pdf

With the EQL the essential object oriented concepts and relational
628.
concepts are integrated in a uniform framework, so that the EQL
supports relational and object-oriented database capabilities. While
the EQL can be used to introduce new abstract data types for ob
jects, handle inheritance, query and manipulate complex ob jects,
and invoke methods of objects, the basis of the EQL is relational in
nature. Objects and values are organized into relations.

Database and Expert Systems Applications.pdf

 5.1 An object oriented prototype In this section we give a short
629.
description of a prototype implementa tion in the Object Oriented
Paradigm. Our approach was to implement each entity type of the
conceptual schema from figure 9 as a SO called class, where a class
consists of a data structure and the methods for accessing the data.

Database and Expert Systems Applications.pdf

630. on specific extensions ~ xE~Y representational concept filing process

’ the designer's model ExER operators !. expression of ~licit data
relationships , I’ and semantic enhancements filing process of
database 1 naming I federative loose database coupling I federation
valne domain restrictions COll'ltraints data data figure 1: application
domain and· WINHEDA bottom up approach Subsequently the
developed three phases approach is dis cussed in more detail.
Chapter 2 describes the user-initiated knowledge integration
approach as a first step to a gradual knowledge integration and the
realization of concepts in a prototype system. Phases two and three
of the approach are described briefly.

Database and Expert Systems Applications.pdf

Yet, since ontologies offer excellent data integration support for

631.
disparate systems, a relational database derived directly from an
ontology can be hypothesized to be the best way to ensure data
integrity and a unique data access point for a large volume of data.
First, modeling a relational database using ontologies (rather than
entity-relationship) allows the reuse of ontology in many tasks
facilitating queries expression as well as ensuring semantic and
structural uniformity [16], [17]. Second, with the axiomatic model
underlying the ontology, data storage and verification can be
automated, increasing data integrity.

Generating_Relational_Database.pdf

632. Figure 1 summarizes the overall configuration of SGML-DB and

STUFF. SG ML-D B contains documents with the following handling
schema: document schema user interface schema transmission
control data.
 Database and Expert Systems Applications.pdf

As shown in Figure 2, the components break down into user

633.
interface, query processor, data access and intelligent retrieval
subsystem. The data access subsystem consists of conventional and
media manager and controls the access to the actual data stored in
relational and media DBMS. The intelligent retrieval subsystem is
composed of parser, gen erator, matcher and description manager.

Database and Expert Systems Applications.pdf

A convenient data structure has to be as compact as possible, and to

634.
support access paths corresponding to different types of
manipulation. A bitmap-compatible structure seems to be an at
tractive solution. Quadtree [16,17] is such a structure, which provides
an interesting technique to code images in a compact way [4].

Database and Expert Systems Applications.pdf

635. If a cooperating XPS is coupled with a DBS and not only data but also
a part of the knowledge is transferred from the XPS to the DBS then
this knowledge must also be available to the other XPS. For this, the
knowledge must be stored and accessable in a suitable fashion. The
solution realized in the Delphi project does not provide these
features, because a part of the transferred knowledge is hidden in
the application program.

Database and Expert Systems Applications.pdf

Operating the KBMS stand-alone this knowledge is beneficial for the

636.
user, too, e. g. if the’ user gets unexpected answers or the resulting
data indicate some regularity the user wants to know whether it is
coincidental. In such cases, it would be necessary to query the
knowledge base and to get knowledge rather than data as result.

Database and Expert Systems Applications.pdf

637. (The intended meaning is /\ AeIT A ""’ VAer A. Sequents consisting

of pre dicates only correspond to clauses by II =} [ ’" VAeIT -,A v
VAer A.) GERALD QUIRCHMAYR Forschungsinstitut fUr
 anwendungsorientierte Wissensverarbeitung ( FAW) Johannes Kepler
Universitat Altenbergstr. 69 A-4040 Linz ( Austria) 732-2468-882
Logical axioms: A =} A Inference rules: Logical rules: L1 (-,- Ieft) : (-,-
right) : L2 ( A-left): ( A-right): L3 (v-left) : (v-right): L4 (""’-left) : b-
rightJ: Structural rules: S 1 weakening ( w-Ieft): S2 exchange
(exchange) : S3 factorization ( f-Ieft) : (f-right) : [=}/l,D -,0, [=}/l 0, [=} /l
[=}/l,-,D C,D, [=}/l CA 0, [=}/l ’ [=}/l, C [=} /l, ° [=},1, CAD C, [=},1 0, [=},1
Cv 0, [=}/l [=}/l, C,D [=},1, CvD [=},1, C D, [=},1 C"",D,[=},1 C, [=},1, °.
[=}/l, C""’D (w-right): [’ is rearrangement of [, /l’ is rearrangement of /l;
A,A’, [=},1 Ao, [0 =},1o [=},1, A,A’ [0 =} ,10, AO’ ° is most general
unifier of A and A’; S4 resolution (res): [=}/l,A A’, II=}A ([, mo =}
(,1,A)o

Database and Expert Systems Applications.pdf

The main problem in building the data model for sequence databases
638.
is how to represent feature descriptions of sequences. A relational
model[2], CYC and interval calculus[9], and a nested relational
model[17] have been tried.

Database and Expert Systems Applications.pdf

2 DATA MODELLING The goal of a data model is to systematically

639.
describe all data that are relevant to the company's business, the
internal structures as well as the linkages to the outside world. This
description is not structured according to a data processing
perspective, but strictly according to the application perspective. This
approach is also termed "function-independent" because data are not
organized with respect to the requirements of particular functions but
rather from a general application point of view that combines the
requirements of all individual functions that may occur within the
enterprise.

Database and Expert Systems Applications.pdf

 Task models represent domain knowledge bases. We repre sent our
640.
task models using the Iris data model whose lin gua franca is the
OSQL query language. Task models are presumed to contain
moderate data volumes but complex structures.

Database and Expert Systems Applications.pdf

Thus, we had to develop a schema which combines traditional

641.
concepts used for information system design ( Le. providing spe cific
data structures) with an adequate work space for diagnostic
procedures (e. g. rules for heuristics).

Database and Expert Systems Applications.pdf

642. Secondly, relational databases present full concep- tual models [16].
Thirdly, they provide a full information resource [16]. Finally, they
offer one of the best techniques for storing and manipulating data.

A novel approach for learning ontology from Relational db.pdf

643. Interview transcripts collected in Amharic language were translated

to English and recorded and coded under the theme category on a
daily basis. Likewise, data captured on field notes from field
observations and ethnography were checked and assimilated to the
relevant theme. In the analysis, primary data and interpretations of
the results, translated direct quote from the respondents were widely
used.

Unemployment Experience of Youth in .pdf

These are divided into rule classes: rules for concepts and their
644.
properties, rules for data types, rules for hierarchical rela- tions,
instances, and axioms. Through these, all database components are
converted into the corresponding ontology components to generate
the final ontology. The approach was implemented using C#.

Ontology learning from relational database a review.pdf

 Abstract—The main purpose of the Semantic Web is driving the
645.
evolution of the current Web by enabling users to find, share, and
combine information more easily. OWL ontologies play a key role in
this effort. It is widely believed that the majority of current Web data
sources are powered by relational databases ( RDB).

OWL Ontology Extraction from Relational.pdf

646. Stanford’s algebraic methodology Jannink and Wiederhold [55]

maintained the ontology using the same methodology as that
employed to generate their ontologies – ontology algebra. When the
knowledge base changes, they use the S operator, which provides a
simple method for assessing the contents of a context to reveal
terms with missing end tags in the data. For instance,
Slen(hw)div20(dictionary) returns the entries of the dictionary,
grouped by length of the head words.

document ontology2.pdf

To clarify this, let us consider how to represent the Establish- Refine

647.
method for hierarchical clas- sification (see earl ier subsection "
Background Work in Knowledge Modeling") using the f ramework
described. Hierarchical classifica- tion is used in many diagnosis sys-
tems as a way o f quickly focusing on possible malfunctions. The
initial state of the classification task is a set of data (e.

Task-Structure Analysis for KM.pdf

[C070] E.F. Codd, A relational model of data for large shared data
648.
banks. In Communications of ACM 13:6, pages 377-387, 1970.

Database and Expert Systems Applications.pdf

The resultant data structure is shown in Fig. 3-1 (a).

649.
2. (one-many) Suppose that we have one-many relationship from
records of type R I to records of type R2. We create, beginning at the
MEMPTR field of each record r of type RI, a pointer cl]l!

Database and Expert Systems Applications.pdf

 Protection against risks like rock bursts, inbursts of wat~r, explosions,
650.
fire, and toxic gases underground plays a CruCIal role in the mining
industry. Moreover the effects of the working itself (horizontal and
vertical dislocations) on the mine and the environment have to be
calculated to avoid or minimize damages. Completeness plays a
crucial role in the information system, since data, which are not
available (in the database), cannot be provided for risk assessment
and precautions.

Database and Expert Systems Applications.pdf

651. Tests of the runtime behavior of the current prototype, using data of a
real mine and involving a large amount of geometric objects and
joins, are promising. The tests also include interaction with
application programs (for risk assessment and correlation of
geological strata) producing graphical output. The results assure that
the infonnation system will be able to manage large amounts of data
and to allow the interaction with additional applications.

Database and Expert Systems Applications.pdf

This would waste a large quantity of data. When joining data of more
652.
editions it would be impossi ble to run the application on affordable
machines. The third one is the different number of relations.

Database and Expert Systems Applications.pdf

(2) Extending relational databases with rules. First order logic is used
653.
as the data model.

Database and Expert Systems Applications.pdf

 The SYLLOG Expert Database System [26] is an expert and data
654.
system shell which makes use of the concept of syllogisms to build
knowledge-bases without the need for programming. The system
uses a near-English language with very few restrictions, executable
speci fications, and improved declarativeness for Prolog-like
programs.

Database and Expert Systems Applications.pdf

Thus, the KB system is required to store and manage potentially

655.
large amounts of sensor data. It is also required to accept this data in
real-time. The engi- neering data may already exist in a database so
the design of the KB system should be flexible enough to allow the
integration of that database.

Database and Expert Systems Applications.pdf

656. The access to the database is realized as a tight coupling [5, 6] that
allows loading data into the expert system at any time during the
inference process. At present state only reading access to the
database is realized.

Database and Expert Systems Applications.pdf

With regard to the intensional query mechanism of KBMS we have to

657.
answer the question, which of the known methods is the best one for
the application [11] and the cQroperation with the knowledge
management. For example, it is quite possible, that the same
functions or structures are used during both the access of knowledge
and the derivation of data.

Database and Expert Systems Applications.pdf

References [1] E. F. Codd. A relational model of data for large shared

658.
data banks. Communications of the ACM, 13(6), June 1970.

Database and Expert Systems Applications.pdf

 2.1 Bottom up and manual One of the aims of this phase is to make
659.
technical prereq uisits for parallel access to the involved multiple
knowledge sources available. The second aim is the preparation of a
high semantical level of representation in order to store explicit
database contents and to express implicit data relationships.
Particularly, this concerns relationships between overlapping
databases, which are autonomous and not coupled so far.

Database and Expert Systems Applications.pdf

The above methods of declustering are heuristics and do not take

660.
into consideration optimal execution of the database operations. The
time spent in exe cuting database operations is well spent if minimal
number of irrelevant tuples are proceesed and to the best of our
knowled~e none of the above methods have taken this criterion into
conaideration on which our methodology is based. The catalog
relations can be used by the query processor to route the database
operations to only those procellOrs which have the data relevant for
that operation.

Database and Expert Systems Applications.pdf

Although there are many databanks of DNAs and proteins to support

661.
research in molecular biology, their integrated database is urgently
requested, because redundancies and gaps among them prevent
effectiveness of such data and knowledge.

Database and Expert Systems Applications.pdf

662. Patient identification Final Tumor diagnosis Figure 3: Structure of the

transfer file Next to generating the structure of the transfer file it is
necessary to convert data types of some attributes, e. g. date of birth.

Database and Expert Systems Applications.pdf

663. Preparation The second phase of the LOFRDB involves data

preparation, which includes the steps to explore and preprocess
(conditioning) data prior. The data exploration consists of checking if
the data sources contain enough semantics for generating ontology
by checking if the RDB contains the complete space of relations and
 the maximum pos- sible combinations of the primary keys and
foreign keys [29]. The second sub-phase is data conditioning, which
refers to the process of cleaning data and normalizing data- sets.

A novel approach for learning ontology from Relational db.pdf

One solution to the problem with a dispersed workforce is experience

664.
repositories. A lot of research has gone into this field, however most
of this research has been focused on large companies and little data
exists on the application of this in small companies [1, 2].

Lecture Notes in Artificial Intelligence.pdf

3 Utilisation of Vocabulary Knowledge Before having the first thought

665.
about the definition of similarity measures, a knowledge engineer has
usually already invested a lot of modelling effort: Prior to settling
similarity measures, the respective application domain—in terms of
attributes and corresponding data types1—must be modelled. This
involves tasks such as determining which case attributes are to be
used, which are indiscrimi- nant and which are possibly dependent
on each other. Moreover, each attribute must be associated with an
appropriate data type and domain, respectively.

Lecture Notes in Artificial Intelligence.pdf

The collected application data can be used for analyzing what is

666.
going on in the fire ground. CAKE retrieves additional information
based on the application data by sending the collected data as
request to the search engines, particularly to the search engines that
are working on guidelines and working instructions. Consequently,
CAKE gets results that match to the application data and contain
guideline or working instruc-

Lecture Notes in Artificial Intelligence.pdf

667. Complementary, relational approaches have much potential for

bringing performance and power to ontology-based data operations.
Finally, more work must be done to decreases the gap between
ontologies and relation databases but the work presented is a step in
this direction.
 Generating_Relational_Database.pdf

: GESANG [7]) - calculating horizontal and vertical dislocation,

668.
especially subsidence damage - topological scanning and risk
protection using knowledge based systems - identification of
geological strata from drilling samples (cf. chapter 6 for more details)
The programs are based on various ~les abo~t deposits, mines,
workings, drilling samples, geologIcal findmgs, and lab data
concerning quality parameters of seams having different formats and
structuring concepts. The basic data, for instance, for calculation of
coal reserves and for constructing the drawing of a mine are similar,
but incompatible. Up to now no database management systems are
used to provide a coordinated information management 87 Two
different types of workstations for markscheiders are in use at several
sites.

Database and Expert Systems Applications.pdf

A Flexible and Extensible Index Manager for Spatial Database

669.
Systems* Hans-Peter KriegeP, Peter Heep2, Stephan Heep3,
Michael Schiwietzl , Ralf Schneiderl 1 Institut fUr Informatik,
Universitlit Miinchen, LeopoldstraBe 11, D-8000 Miinchen 40 2 NANU
NANA GmbH, Lange StraBe 48, D-2900 Oldenburg 3 Praktische
Informatik, Universitiit Bremen, BibliothekstraBe, D-2800 Bremen 33
AbslIaCt The management of spatial data in applications such as
graphics and image processing, geography. as well as computer
aided design ( CAD) imposes stringent new requirements on socalled
spatial database systems. In this paper we propose a flexible and
extensible index manager for efficient query processing in spatial
database systems by integrating spatial access methods.

Database and Expert Systems Applications.pdf

670. RECURSION WITH THE GRAPHICAL QUERY LANGUAGE CANDID M.

Schneider C. Trepied Laboratoire Infonnatique Universite de
Clennont-Ferrand II 63177 Aubiere Cedex ( France) ABS1RACT
CANDID is a graphical interface for data base manipulation desi gned
for the final user. It is based on a generic semantic model using
attributes and constructors. Each schema is represented by a graph
 where the nodes are associated to different object types and where
the edges represent relationships between types.

Database and Expert Systems Applications.pdf

671. 3 The expert system The IRIDA expert system has a twofold
purpose: during data input the expert system is dedicated to
automatic document indexing using text analysis and context
analysis techniques; at the operational stage the same knowledge
base can be used for intelligent information retrieval, allowing a list of
topics expressed in pre-coordinated controlled language
(classification) to be matched with a list of potential, subsidiary topics
contained within each class. .

Database and Expert Systems Applications.pdf

For each T ET RT∈ ∪ , there are ( 0)n n ≥ foreign keys ( FKs). If 1n ≥

672.
, then the n foreign key(s) are ( ) ( ) ( ) ( )1 2, , , nfk T fk T fk T attr T⊆
, where each value of the attribute(s) in ( ), 1, ,ifk T i n= references
the relevant value of the attribute(s) in the primary key ( )pk R of
another entity table R ET∈ , denoted as ( ) ( )( , ) iRef fk T pk R .

OWL Ontology Extraction from Relational.pdf

The second algorithm attempts to con- struct the ontology of

673.
individuals ( ABOX) using data from different RDBs records based on
the ontological model. The proposed method can do the conversion
in OWL of all the types of tables together with their columns in an
original way because it keeps a trace of the original nature of the
relational database tables by creating the two superclasses Entity
and Association. The paper provides no rules for map- ping not null
or unique constraints in the case of foreign key attributes, which can
be an important shortcoming.

Ontology learning from relational database a review.pdf

The Student Database Within the polytechnic there is a large

674.
database which holds information on prospective students who are
applying for undergraduate courses. This system is used by all the
 schools in the polytechnic. Each of which applies its own admission
rules to the data held on the DB about each student.

Database and Expert Systems Applications.pdf

675. Given these facts, it seems natural to look for a link between the two
different approaches to data modeling. In this paper we have tried to
find ways of translating some features (i. e., inheritance, or
generalization, and object sharing) of object oriented databases into
(nested) relational databases.

Database and Expert Systems Applications.pdf

Hence, partial or approximate match between descriptions of

676.
multimedia data and user queries is generally required during
multimedia data retrieval. We propose an intelligent approach to
approximate match by integrating both object-oriented and natural
language understanding techniques. In order to make the query
specification process easier we also develop a graphical user inter
face supporting incremental query specification and a natural way of
expressing joins.

Database and Expert Systems Applications.pdf

Prolog system coordinator t t Compiler WAM CLA I I t t CSM I CBM

677.
Communication modules t Oata storage system -+- control flow Fi
ure 1: The Prolo s stem architecture The Prolog system consists of
the following subsystems: a com piler to compile arbitrary Prolog
programs into code clusters containing Warren Abstract Machine
code, a Warren Abstract Machine ( WAM) and an external data
storage system ( OSS), e. g. a file system, database system, etc. to
store und retrieve 155 code clusters. In order to keep the WAM and
compiler indepen dent of specific OSS details, a cluster storage
module ( CSM) and a cluster buffer manager ( CBM) have been
designed as inter faces between the Prolog system and the data
storage system.

Database and Expert Systems Applications.pdf

 First, we allow multimedia data in the form fields. Second, we
678.
consider a form as a complex object inside the database [5], [2], [16],
[21]. Third, because of the interactive nature of a form dis played on
the screen, actions can be performed upon the form, letting the
system react accordingly.

Database and Expert Systems Applications.pdf

679. With regard to databases as data storage systems, contri butions in

the field of semantic integration and translation of database schemes
is one of the topics in research activi ties on Federative Database
Systems ( FDBS), as it is called. A FDBS consists of distributed,
possibly heterogeneous and autonomous component databases, that
are part of a fed eration by allowing partial and controlled sharing of
their data.

Database and Expert Systems Applications.pdf

Set T consists of transactions TI , T2 , ’" , T m and a fictitious

680.
transaction To called the initial transaction. This transaction is a write-
only transaction which writes the initial values of all data items of the
DB. Each transaction Ti E T, i E (1, m), is a poset Ti = ( Ti, <i), where
Ti is a set of database operations issued by Ti and <i is a partial’
order relation which specifies the execution order of operations in Ti.

Database and Expert Systems Applications.pdf

The services rendered by the system include browsing, data re

681.
trieval, information retrieval, and the combination of data and
information retrieval. The non-linear structure of hypermedia
information is particularly suited for browsing activities. Although
browsing appears to be an unsystematic way of searching for
information, our belief is that many user needs may be satisfied when
a flexible browsing tool is available.

Database and Expert Systems Applications.pdf

The query languages proposed in [5] and [13] provide ranked output
682.
by means of the language constructs RANK_BY and RANK
respectively. These extended query languages do not allow
 arithmetic computations on the data that has been retrieved by
imprecise search criteria. The query language VAGUE [11] contains a
SIMILAR-TO comparator which is determined by different metrics.

Database and Expert Systems Applications.pdf

4.2.1 Conceptual Differentiation The heterogeneous information

683.
described with the model requires a powerful and flexible model
structure. Thus, the degree of differentiation will generally be higher
than in a regular enterprise-wide data model. Because of the
structural diversity of the information to be captured in the structure,
information carriers will move in status from attribute to entity type.

Database and Expert Systems Applications.pdf

496 It should be stated, that, in addition, the data modelling approach

684.
outlined above requires a 8Ophist~cated functionality to ensure
consistency of the data base ill the stage of model implementation.
The task is in translating the contents of the individual data bases
according to the rules described by the organization-wide model and
its links to the individual models. Those rules are described by the
generalization operators used as links between the individual models
and the overall organizational model; they go beyond the integrity
rules usually stated for a data base [17].

Database and Expert Systems Applications.pdf

Here we provide the same uniform query lan guage, OSQL, both for
685.
the shared database, the extraction of data into work areas, and for
representing the work areas themselves.

Database and Expert Systems Applications.pdf

686. It may be possible to refer to an expert in the studied domain to

resolve some problems concerning the gathered data such as the
database conceptualization problems [47].

A novel approach for learning ontology from Relational db.pdf

 So the system was designed in a generic way so that appropriate
687.
features (together with their acquisition and exploitation strategies)
can be added to the schema without affecting the core functionality or
the interfaces of the involved services. The core of this generic
infrastructure is a data model for context information based on RDFS
that is capable of representing imperfection and dynamic phenomena
like aging as well as the context history [6] for all context features.

Lecture Notes in Artificial Intelligence.pdf

Use Case Outputs Recording of data of the new user. Criterion of

688.
Acceptance Data of the new user are stored in the database of the
Local Branch. Related Expectations Database management system.

Lecture Notes in Artificial Intelligence.pdf

html; an open multi-agent system ( MAS)) and InfoSleuth. The second

689.
is based on the concept of mediators [35] that provide intermediary
services by linking data resources and application programs.
Examples include the TSMISS at Stanford University.

document ontology2.pdf

690. Output: The resulting OWL DL Ontology ( )( ) ,o oO S ID Axiomφ= =

that is defined by a translation φ . 1 The translation from relational
symbols to OWL class and data range identifiers: 1.1 for each
entity/relationship table symbol T ET RT∈ ∪ , create a class identifier
( ) oT CTIDφ ∈ ; 1.2 for each datatype symbol d DT∈ , create a data
range identifier ( ) od DTIDφ ∈ by mapping it to an XML Schema
datatype reference; 2 The translation from relational elements to OWL
class, datatype/object property identifiers, and class/property axioms:
2.1 for each table T ET RT∈ ∪ and its non-foreign key attribute set ( )
1 1[ : , , : ], 1h hNonFkSet T A d A d h= ≥ , 2.1.1 create datatype
property identifiers 1( ), , ( )h oA A DPIDφ φ ∈ and property axioms:
DatatypeProperty( 1( ) Aφ domain( ( ) Tφ ) range( 1( )dφ )); …
DatatypeProperty( ( ) hAφ domain( ( ) Tφ ) range( ( )hdφ )); 2.1.2
create a class axiom for the table: Class( ( ) Tφ partial restriction( 1( )
Aφ allValuesFrom( 1( )dφ ) cardinality(1)) … restriction( ( ) hAφ
allValuesFrom( ( )hdφ ) cardinality(1))); 2.2 for each foreign key
attribute ( ), , 1, , ( 1) iA fk T T RT i n n∈ ∈ = ≥ such that ( ( ), ( )), iRef
 fk T pk R R ET∈ , 2.2.1 create an object property identifier ( ) oA
OPIDφ ∈ ; 2.2.2 create a class axiom: Class( ( ) Rφ partial restriction(
( ) Aφ allValuesFrom( ( ) Tφ ))); if ( ) ooRelTab T true= then add
maxCardinality(1) to ( ) Aφ ’s restriction else if ( ) omRelTab T true=
and ( )A Tpk= then add maxCardinality(1) to ( ) Aφ ’s restriction; 2.2.3
create a property axiom: ObjectProperty( ( ) Aφ domain( ( ) Rφ )
range( ( ) Tφ )); 2.3 for each foreign key attribute ( ), , ( ) , 1, , ( 1) iA
fk T T ET subEntityTab T false i n n∈ ∈ = = ≥ such that ( ( ), ( )), , iRef
fk T pk R R ET R T∈ ≠ and A references the value of attribute ( )B pk
R∈ , 2.3.1 create a class identifier ( ) _ ( ) oT R CTIDφ φ ∈ and two
object property identifiers ( ), ( ) oA B OPIDφ φ ∈ ; 2.3.2 create two
class axioms: Class( ( ) Tφ partial restriction( ( ) Aφ allValuesFrom( (
) _ ( )T Rφ φ ))); Class( ( ) Rφ partial restriction( ( ) Bφ
allValuesFrom( ( ) _ ( )T Rφ φ ))); if ( ) omEntityTab T true= or ( )
weakEntityTab T true= then add maxCardinality(1) to ( ) Aφ ’s
restriction and minCardinality(1) to ( ) Bφ ’s restriction else if ( )
ooEntityTab T true= then add maxCardinality(1) to ( ) Aφ ’s restriction
and minCardinality(1), maxCardinality(1) to. ( ) Bφ ’s restriction; 2.3.3
create two property axioms: ObjectProperty( ( ) Aφ domain( ( ) Tφ )
range( ( ) _ ( )T Rφ φ )); ObjectProperty( ( ) Bφ domain( ( ) Rφ )
range( ( ) _ ( )T Rφ φ )); 2.4 for each foreign key attribute 1( ), ( )A fk
T subEntityTab T true∈ = such that 1 1 1( ) ( ) [ : , , : ], 1k kfk T pk T A
d A d k= = ≥ , 2.4.1 create datatype property identifiers 1( ), , ( )k oA A
DPIDφ φ ∈ ; 2.4.2 create a class axiom: Class( ( ) Tφ partial
restriction( 1( ) Aφ allValuesFrom( 1( )dφ ) cardinality(1)) …
restriction( ( ) kAφ allValuesFrom( ( )kdφ ) cardinality(1))); 2.4.3
create property axioms: DatatypeProperty( 1( ) Aφ domain( ( ) Tφ )
range( 1( )dφ )); … DatatypeProperty( ( ) kAφ domain( ( ) Tφ ) range(
( )kdφ )); 2.5 for each 1 2, , , , , 1mT T T T ET m∈ ≥ such that 1( ) ( )
msubEntityTab T subEntityTab T true= = = , ( ) superEntityTab T
true= and 1( , ) ( , ) msubOf T T subOf T T true= = = , 2.5.1 for 1..

OWL Ontology Extraction from Relational.pdf

691. Frame: Op,= Block: Fr = (OF" WF,) where { FrTid, AI’ A 2 , , Ap ’

FLaOb, FCoOb } Bl = ( OBI’ wB/) where OBI = {B1 Tid, AI’ A 2, , Ap ’
F LoOb’ FCoOb } The F ~ composite attributes refer to inherited
attributes from the previous four basic classes (on the hierarchical
structure). From this point on, the hierarchy of the ODA data model
may change according to user specifications. There are two possible
ways of doing this: by means of the’ Document Application Profile (
DAP) and by means of direct user specifications.
 Database and Expert Systems Applications.pdf

SCIDKORRE - Correlation of geological strata Starting from drilling

692.
samples this expert system identifIes geological strata, especially
seams. It analyses material properties of the samples, determines
similarities with respect to thickness and quality parameters resulting
in a correlation between the different samples. In order to integrate
these results with existing geological data access and transfer to a
database system is necessary.

Database and Expert Systems Applications.pdf

The paper is organized as follows: Section 2 discusses re lated work.

693.
Section 3 addresses fundamental problems and outlines the
architecture of the Multimedia Database Man agement System (
MDBMS). Section 4 describes the natural language interpretation
capabilities of the parser, Section 5 the approximate match algorithm
used for retrieval of mul timedia data and Section 6 gives a short
overview of the user interface.

Database and Expert Systems Applications.pdf

694. In section 2, we present a brief overview of the SUPER environment

model and its main features. In section 3, we present the visual
interfaces currently specified and developed. the schema editor and
the query editor. Section 4 shows why and how we used toolboxes to
implement the ERC+ data model and to support the direct
manipulation paradigm of the interfaces.

Database and Expert Systems Applications.pdf

695. The schema and query editor have been analyzed, with a focus on
functionalities and the underlying design choices. Finally, we have
shown the suitability of toolboxes to implement the data models and
to support the direct manipUlation plll'adigm.

Database and Expert Systems Applications.pdf

In this paper, we also discuss several con cepts in object-oriented

696.
databases, and propose an object onented data model which, we
 hope, will serve as basis for an object-oriented database system
incorporating deductive capabilities.

Database and Expert Systems Applications.pdf

697. The Data Management Module ( DMM) main tains the database and
provides access to that database to the other modules in the system.
The Database ( DB) stores static information about the device and its
compo nents, a history of the faults that have occurred, and the
sensor data passed from the Data Analysis Module. A complete
description of the Data Management Module is given in Section 4.

Database and Expert Systems Applications.pdf

698. O. Introduction In the Relational Data Bases context [6,11, 12], the
view concept has been developed to show data to the user. In fact,
view definitions allow to show derived data, to hide undesirable data
and so on, by means of classical operations of the relational algebra.

Database and Expert Systems Applications.pdf

699. Secondly, the organization needed a way to examine how and why
individual projects adapted their base process, and how successfully.
Without this kind of explicit knowledge supported with qualitative and
quantitative data, the process changes in organizational level would
be a shot in the dark rather than validated learning to be shared and
diffused organizationally.

Lecture Notes in Artificial Intelligence.pdf

700. However, in this paper we emphasize the data and problem

understanding perspective. We see the main contribution of this
paper in demonstrating how the many-sided analysis of a real-world
problem has been performed and how the application of diversified
DM techniques has helped us to enhance our understanding of the
nature of the data that represents the problem.

Lecture Notes in Artificial Intelligence.pdf

 A DEDUCTIVE DATABASE SYSTEM WITH APPLICATIONS TO ROUTE
701.
PLANNING Malcolm C. Taylor Department of Computer Science
University of Houston Houston, TX 77204 ABSTRACT The issue of
deductive query optimization has received widespread attention, yet
the problem of dealing with functions defined on com plex data
structures remains unresolved. In this paper we present an approach
which allows for abstract data types and arbitrary user defined
functions within the framework of an extended relational model.
Functions may appear in both rules and queries, and the op
timization process uses a combination of term rewriting and distribu
tion of selections.

Database and Expert Systems Applications.pdf

702. An important application area for active mediators is in the support of

systems where persistent data is checked out into main memory
work area8, which are here represented as ac tive task models.
During long sessions the user would work directly with specific and
hence efficient active task mod els and only occasionally check data
back into the central database. This complements the view object
paradigm pro posed by Wiederhold [22] and implemented in the
PENGUIN project [2].

Database and Expert Systems Applications.pdf

Let be the following composite axiom A OP qt (B ∩ C) where A, B,

703.
and C are classes, OP is an object property, and qt is a quantifier:
some, only, min, max, exactly; it is possible to express it in 3 simple
axioms: A OP qt Z ; Z isa B ∧ Z isa C where Z is a new class subclass
of Thing. In a relational database, such operators are defined in the
conceptual model and may have different transformations [3]
generating different structures and constraints. This may be an issue
when using popular medical and biomedical ontologies (see
ontologies in OBO foundry repositories [24]) because composite
axioms are frequently used. There is a need for a conversion process
that uniformly converts such axioms by preserving the complete
semantic.

Generating_Relational_Database.pdf
 One of the serious problems by retrieving code of Prolog know ledge
704.
bases from secondary storage is the large load time, which is quite
prohibitive for a run time retrieval of large knowlegde bases. The load
time of knowledge bases comprises the transfer time and the
translation time; the latter incures 90% of the total load time [ Boc90].
The following diagrams show an exponential load time by an
increasing knowledge base size; the load times are measured by
interpreting (consult) and compiling (compile) source code by
QUINTUS Prolog 2.4 under AIX on a IBM RT computer: ·The work
reported here has been carried out at IBM Stuttgart as part of the
EUREKA project PROTOS ( EUS6): Prolog 1’2.

Database and Expert Systems Applications.pdf

705. Fig. 3: Extended System Architecture The data model FGD (

Functional Global Data model) [7] represents a functional interface,
which supports abstraction mechanisms of recent programming
languages and semantic data models. The data manipulation
language of FGD is based on functional languages like Galileo [8, 9,
10] and is ex tended for the integration of especially
pre:.existing heteroge neous databases and for accessing the
data in a transparent way as described above.

Database and Expert Systems Applications.pdf

It was already mentioned that queries are made from the viewpoint of
706.
the current state. The objects in the result sets must therefore be
available in the current state, because only for current objects the cor
responding functions and predicates are defined. Information about
deleted objects can only be represented by data values because they
are available for the whole database lifetime.

Database and Expert Systems Applications.pdf

Object oriented database is currently an active research area. This is

707.
because object oriented concepts can more easily model the real
world and capture its semantics than more tradi tional DBMSs.
Besides this, the object oriented data model allows us to reuse
components by producing enhancements based on the original.
 Database and Expert Systems Applications.pdf

The designers would work con currently on the design in the active
708.
task model using some supporting software tools. The tools would
use monitors to notify designers when some other designer modifies
some data that is of interest. Winslett et al [27] describe an ar
chitecture for consistency maintenance in a design database, which
could be supported by active task models.

Database and Expert Systems Applications.pdf

The knowledge which has not been covered by the class definition
709.
has been represented by Prolog-clauses. Prolog-DB processes
meta-rules with data from relations and results from database
queries.

Database and Expert Systems Applications.pdf

6.1 FROM THE ONTOLOGY TO A FORMAL MARKUP LANGUAGE As it

710.
seemed appropriate to translate the business model ontology into a
formal description language and as the eXtensible Markup Language
XML has rapidly become the first choice for defining document and
data interchange formats I have chosen this technology to formalize
the business model ontology.

The_Business_Model_Ontology_a_propositio.pdf

Basically there is always the question where ontologies have benefits

711.
over conven- tional approaches like established ERP-Software with
relational data models. Yes – it is true that everything what we can do
with ontologies can be done with some other technology as well. But
there is always the question how much effort do you have to pay as
invest for your solution in contrast to often very elegant ontology-
solutions with lots of implicit benefits like:

Lecture Notes in Artificial Intelligence.pdf

712. Another type of related work concerns the RDB-to- RDF mapping
approaches or tools, which focus on data mapping from relational
databases to RDF datasets. Such mappings provide the ability to
 view existing relational data in the RDF data model. These tools, such
as D2RQ [21], Virtuoso RDF View [22], and Triplify [23], offer a virtual
SPARQL endpoint over the mapped relational data, or generate RDF
dumps, or offer a Linked Data publishing interface.

OWL Ontology Extraction from Relational.pdf

Breakthroughs can come from either completely different settings or

713.
classes of systems particularly adapted to specific applications. We
can seek for such improvements from recovering background knowl-
edge (§9), for example, from the linked open data cloud as it
represents a large and continuously grow- ing source of knowledge.
Another source of quality gains is expected from the working
environment in which matching is performed.

Ontology matching state of the art and futur.pdf

714. 1 2( ) ( )( ) fk T fk Tpk T ∪= then { ( ) mmRelTab T true← ; continue};

if T has n foreign keys and one primary key ( )pk T such that 1 2( ) ( )(
) , 3nfk T fk T fkpk T n∪ ∪ ∪= ≥ then { ( ) naryRelTab T true← ;
continue}; if T has exactly one foreign key 1( )fk T and one primary
key ( )pk T such that 1 1( ) ( ), ( ( ), ( )),fk T pk T Ref fk T pk R T R= ≠
then { ( ) subEntityTab T true← ; ( ) superEntityTab R true← ; ( , )
subOf T R true← ; continue}; if T has. exactly one foreign key 1( )fk T
and one primary key ( )pk T such that 1 1( ) ( ) , ( ( ), ( )),fk T pk T Ref
fk T pk R T R∩ = ∅ ≠ then if many values in ( ( )) Data pk T relate to
one value in 1( ( )) Data fk T then { ( ) omEntityTab T true← ;
continue}

OWL Ontology Extraction from Relational.pdf

[ MA77] AMakinouchi. A consideration of normal form of not-

715.
necessarily normalized relations in the relational data model. In Inti.

Database and Expert Systems Applications.pdf

1 INTRODUCTION One important question in the trend to more open

716.
office systems is whether design methodologies and modeling
methods for such systems will be different to those now used in
practice. Many conventional design methodologies are oriented
towards constructing systems with a fixed structure of both data and
 function and do not support semantics found in open operations.
Thus the option for supporting open system design by extending
conventional methodologies is not easy given their emphasis on data
structure and functionality but no semantics to represent user
coordination and responsibilities of roles within system activities.

Database and Expert Systems Applications.pdf

717. DS can handle recursive rules, but it has not been used with abstract
data types and arbitrary user-defined functions. Although DS in its
basic form has been shown to be less effective than magic sets [13],
it can be made equivalent by introduc ing a rule rewriting phase (in
the style of the magic sets method) be fore translating into relational
algebra.

Database and Expert Systems Applications.pdf

718. Thus, this proceeding assures the highest possible degree of

reliability in the basic data for the various design, prediction and
analysis tasks. Let us illustrate the importance of completeness and
reliability by a simple example: If a runway is planned to be advanced
at point x, the miner underground has no possibility to recognize, that
behind the face (very near point x) there is a water zone with a height
of 200 meters. If the database would not include this information,
advancing at point x would immediately cause the bursting of the
water at a pressure of 20 attn.

Database and Expert Systems Applications.pdf

719. A Friendly and Intelligent Approach to Data Retrieval in a Multimedia

DBMS Daniel A. Keirn!, Kyung-Chang Kim, Vincent Lum Department
of Computer Science2 Naval Postgraduate School Monterey, CA
93943 - 5100 Abstract Manipulation of multimedia data is not
straightforward as in conventional databases. One main problem is
the retrieval of mul timedia data from the database with the need to
match the contents of multimedia data to a user query.

Database and Expert Systems Applications.pdf

Introduction A multimedia database management system supports

720.
the management of multimedia data, which includes image and
 sound among others, in addition to supporting conventional
databases. Multimedia systems are currently gaining a lot of attention
because technology today has made it possible to capture and store
multimedia data in computers. Many ap plications like military,
publishing or instructional routinely need multimedia data.

Database and Expert Systems Applications.pdf

721. One important achievement of the MDBMS system is the efficient

method for the retrieval of multimedia data by way of inexact
matching. In conventional databases, retrieval of standard numerical
and alphanumeric data is handled by utilizing the content of the data.
The fundamental problem that one must face in the context of a
multimedia database is the question of how to provide content
search.

Database and Expert Systems Applications.pdf

The algorithm We now describe the algorithm for a data parallel

722.
computa tion of the predicates of a recursive clique. In a first phase
the rule system is analyzed in the main program and the database
operations of the tasks are determined according to the recursive
rules. Input are the rules of the recursive clique and the query
predicate with its instantiations.

Database and Expert Systems Applications.pdf

Further problems to be considered are, for example, efficient data

723.
butTering in the expert system, communication and database update
opera tions.

Database and Expert Systems Applications.pdf

724. The last database operation of the input sequence, i. e. T1 : r(y),

reads the version y5· The multi version schedule produced by M VTO
algorithm complemented by the dop protocol is the following: To:
w(x,y) To: c Tl : rex) T2 : r(x,y) T2 : w(x,y) T 2 : c T3: r(x,y) T3: w(x,y)
T3: c Tl : r(y) Rdmv = {(To:w(x), T1 :r(x)),(To :w(x), T2:r(x)), (To: w(y),
T2: r(y)), ( T2 :w(x), T3: rex )), ( T2: w(y), T3 :r(y)), (To :w(y), T1 :
r(y))} 7 Final conclusions In this paper we have presented the novel
dynamic overwrite pro tocot (dop) which compared to conventional
 natural overwrite protocol, used in practice, minimizes the number of
transaction abortions caused by a bounded amount of space for
holding data item versions. The main idea of this new overwrite
protocol lies on the finding sets of data items versions which can be
safely overwritten because no active transaction needs to access.

Database and Expert Systems Applications.pdf

725. Revealed Processes in Knowledge Management 447 recording the

user name, it is not shown in Table 1. The events table records
information about events — for the event data itself, the system uses
the body table, a portion of which is shown in Table 2.

Lecture Notes in Artificial Intelligence.pdf

Finally, the Transformation Server delivers the extracted, aggregated

726.
information into the desired formats to other Business Intelligence
systems such as SAP Business Information Warehouse or Microsoft
Analysis Server. Also, the Transformation Server interactively
communicates with these systems using various interfaces, such as
special database formats, XML messaging, and Web services.

Lecture Notes in Artificial Intelligence.pdf

3 Conclusion In this paper, we have defined a semantic database

727.
language which incorporates both the advances of semantic data
models and that of database pro gramming languages. Our goal was
to take the best of the two worlds while keeping the performance and
feasibility of databases. The major points of our proposal are the
following: (1) Serendip incor porates what we think are the most
important and useful features of semantic data models: abstrac tion,
object identity, ISA relationship, complex structures and a declarative
definition of integrity constraints; (2) Serendip is based on the notions
of abstract data types and views.

Database and Expert Systems Applications.pdf

Summary A major problem faced in a multimedia database system is

728.
the retrieval of multimedia data such as a sound or an im age. Media
data is intrinsically rich in semantics and con- ventional search
methods used in databases and information retrieval systems may
 not work or are of little use. Most re search on intelligent IR systems
are concerned with natural language processing and deductive
capabilities based on extended semantic models of document
content.

Database and Expert Systems Applications.pdf

In this paper, we have chosen the object-oriented data model given

729.
in [4] since it not only has the properties of an OODBS (versions [13],
CQmposite objects [4], dynamic schema evolution [5], and
multimedia data management [25]), but also supports a formal query
model. Figure 1 shows the schema for a computer database. In this
figure, circles represent classes, thin edges denote the is-a
relationship, while thick edges represent the is-part-oJ relationship.

Database and Expert Systems Applications.pdf

730. Network trend analysis for capacity or topology management

requires the accessibility of historical performance data. For this
reason, among others, the control unit supports database functions
to store network performance data after it has been processed or
reduced to relevant information for long term usage.

Database and Expert Systems Applications.pdf

In our MDBMS system we use the approach of content based search

731.
by means of verbal descriptions on the con tents of multimedia data.
We argue that the well known key word approach to content
description is not suitable because it has been known to be imprecise
and the users often have difficulty in focusing the search to data of
interest. Hence, we adopt the natural language approach to content
descrip tion as a more viable option.

Database and Expert Systems Applications.pdf

732. This application is simi lar to EIS about properties of chemical

substance that IIct I\A pollutants (see e. g. [31]). In this project, 15
search dialo! l:s with existing materials data systems were observed,
where 14 of the dialogs did not give satisfactory results for the user.
 Database and Expert Systems Applications.pdf

g. the functional data model [ Ship 81], POSTGRES [ SR 86] or the

733.
EXODUS proposal [ CDRS 86] and others are direcdy concerned with
the modeling of spatial data like [ Guet 89]. In the following we do not
propose a new data model but we would like to describe those
modeling capabilities which every spatial database system should
support in our oppinion in any case. Furthermore, we will emphasize
that our index manager may be integrated in a straightforward way
into systems which provide those modeling capabilities.

Database and Expert Systems Applications.pdf

734. The remainder of this paper is organized as follows. In Section 2, we

review the ex tended relational data model and an algebra over the
model. The operators of the algebra permit to combine data from
partially compat ible relations and are thus useful in resolving relat.

Database and Expert Systems Applications.pdf

2.1 Merits of Object-oriented expression£91 In a system that handles

735.
various media, it is necessary to judge the media types and the
processing procedures to be applied, because the procedures are
different for each media, even if they have the same processing
purpose. Thus, multimedia information is formed into one unit having
a series of labels, such as data and procedures, and is stored and
managed as a single object. This object capsulation propeny enables
the responsibility for selecting the procedure that fits the media type
of the object to be transponed from the application side of media
service to the supply side.

Database and Expert Systems Applications.pdf

 1.2 Query Languages Query Languages are usually closely related
736.
to the underlying data model. For the relational model two kinds of
query formalisms can be distinguished. The first one consists of the
class of algebra approaches, whicll allow to select information in an
operational Way by use of a couple of basic operations.

Database and Expert Systems Applications.pdf

737. Before starting the conversion, the procedure creates two classes,
Entity and Association, which are superclasses of all the classes that
will be created. The tables representing entity inherit from the class
Entity, while the tables representing many-to-many relationships
inherit from the class Associa- tion. The algorithm creates various
classes with its data and object properties, considering various
patterns of elements in the RDB schema.

Ontology learning from relational database a review.pdf

In the database literature several models have been proposed for

738.
representing roles during the data modeling process. In our view the
most important contribution to the field of role model ing is
exemplified by the seminal work of Bachman on the role data model
[6]. The definition of the role concept in the role model is taken from
the theatrical context and is used to mean a behavioral pattern which
may be assumed by modeled enti ties in a problem domain.

Database and Expert Systems Applications.pdf

2.3 SUbtyping The is-a relationship is a central primitive in most

739.
semantic data models. In thO( Serendip language, this relationship is
expressed in terms of sub typing. The subtyping relationship mainly
corresponds to the inheritance of object-oriented systems.

Database and Expert Systems Applications.pdf

740. Recognizing the success of the relational model, many researchers

have sought to extend that model to include support for more com
plex data structures. One attraction of an extended relational model
is that existing ideas on query optimization can be adapted to work
 on the more complex data structures [15]. Deductive rules can also
be handled in this framework.

Database and Expert Systems Applications.pdf

741. Current VLSI/ CAD systems generally use a file system provided by
an operating system to store design data. Although these systems
show good performance, they do not achieve a level of integration
that accrues from a centralized database management system. For
example, to use rnl, which is a timing logic simulator described in the
VLSI design tools reference manual [1], one needs to create a
network description .

Database and Expert Systems Applications.pdf

The methodology we adopt consists of associating natural language

742.
captions to each multimedia data and using the description to
retrieve the rel evant data. More precisely, the description of a
multimedia data is matched against the description of a user query
which is also expressed using natural language captions. The major
problem with this approach is that it is generally the case that the
description of a multimedia data does not exactly match the
description of a user query.

Database and Expert Systems Applications.pdf

743. The software that we developed addresses the efficiency issue with
an architecture containing a master module, which create
subprocesses that will read the data and send it to the master
through interprocesses communication channels. Each subprocess
(three of them presently exist) is designed to optimize the access
time to the particular data it must handle. The flexibility problem was
resolved by defining a conceptual view of the data, where each
record is described in a similar way, independent of its actual
representation in the database.

Database and Expert Systems Applications.pdf

744. Automatic summarization of map The automatic creation method for

a summarized map and drawing based on the semantic model of
multimedia is described here. The proposed automatic
 summarization method is different from the road guide map
generated by conventional media-oriented processing or the macro-
map produced by the thinning out procedure. The creation method
has a feature for acquiring the method of processing data which is
necessary for completing the media structure from the user deftned
semantic structure.

Database and Expert Systems Applications.pdf

This means that some write oper ations must overwrite old versions
745.
of data items in order to create new ones. This paper presents the
novel dy namic overwrite protocol which compared to conventional
natural overwrite protocol used in practice minimize the number of
transactions abortions caused by the limited storage space. The main
idea of this new overwrite proto col lies on finding data item versions
which can be safely overwritten because no active transaction
accesses them in the future.

Database and Expert Systems Applications.pdf

746. In this paper we have presented the results of our experimental study
in AR prediction for nosocomial infections. We have achieved rather
high generalization accuracy (84.5%) that is quite promising in terms
of better understanding the problem and patterns of AR. The results
were achieved using data with patients having meningitis over the
last three years only, and we plan to continue our analysis of the
whole NSI database of nosocomial infections including older data
collected since 1997.

Lecture Notes in Artificial Intelligence.pdf

Consequently, the context is an information container for any kinds of

747.
objects used by the workflow. The context comprises administrative
data, workflow control data, workflow relevant data, and application
data that have been defined by Maus [6]. Due to the possibility of
nested (sub)workflows, local con- texts can be nested as well by
following the concept of inheritance.

Lecture Notes in Artificial Intelligence.pdf

 Another important problem to which the de signer of an OODBS is
748.
faced, is to define a design methodology which would be consistent
with the object-oriented approach. Up to now, there exists a lot of
proposals for database design. They range from the Abrial
proposal[2] to sophisticated seman tic data models such as IFO[l] or
INSYDE[13].

Database and Expert Systems Applications.pdf

Matching In this chapter, we propose new ways of matching natural

749.
language descriptions of multimedia data with query speci fications.
The key to our matching process is the use of do main knowledge
represented using the notion of class hierarchy borrowed from the
object-oriented field.

Database and Expert Systems Applications.pdf

750. In IQL [ AbKa 89], in addition to the concept of values, relations are
part of the data model. A relation is actually a set value with ~
identifier, which makes it accessible, and relational ? perat10ns _ are
easily applicable.

Database and Expert Systems Applications.pdf

3.1 Access to External Databases Mine planning operations are

751.
based on a large amount of technical, economical, geological and
spatial data, which are distributed to several autonomous databases.
These databas es are managed by different database management
systems and based on various data models.

Database and Expert Systems Applications.pdf

752. The programmer must specify what happens when the value of a
monitored view changes. A convenient way to do this is to specify a
tracking procedure or tracker, which is a proce dure of the application
that is invoked by the DBMS when monitored data change. The DBMS
thus keeps track of which tracking procedures monitor which object
attributes and contains a mechanism to call the tracking procedures
upon data changes.
 Database and Expert Systems Applications.pdf

Labour market assessment – Market trends and opportunities in

753.
Ethiopia and the Gulf II. RESEARCH APPROACH 17 When reconciling
the data and selecting the most important findings, the triangulation
logic shown in table 4 was applied.

Labour market ASSESwcms.pdf

To achieve this upgradeability, we need to decouple a document’s

754.
data from the standards used to represent the data. We argue that
this separation between data and form can be achieved using a
technique called compilable templates, which was recently
developed by the authors in a Personal Health Record ( PHR) pro-
ject. Web services offer a compelling means of implementing such
templates owing to their language and platform neutrality.

Lecture Notes in Artificial Intelligence.pdf

755. The problem with many of the methods mentioned is that they have
remained at the prototype stage and are not available for use by the
community. In fact, some of these methods have stopped at a
development stage and are not yet fully- fledged products. In other
cases, these approaches have not yet been applied to real-world
databases to verify their performance in automatically converting
relational data- bases into ontologies.

Ontology learning from relational database a review.pdf

 Our requirement-based re-engineering approach differs from ([
756.
P&B94], [ HDA87], [ M&M90], [ M&S89], [ NEK94], [ MNBBK94],[
HCTJ93]) that we couple our extended entity-relationship modeling
methodology and tools with model integration processes to produce
a normalized data model. In a re-engineered data model, the
physical data objects, pertaining to the physical schemas, link to the
relevant logical data objects, external data objects, business
requirements, and system rules.

A Requirement-Based Approach to Data Modeling and Re-

engineering.PDF

This was also described in [2] as a " Homogeneous Approach". The

757.
second method of integration is t~e e~hanced ES approach. In this
approach the expert system IS gIVen the data management functions
of a database.

Database and Expert Systems Applications.pdf

Our interfaces avoid using different modes of operation, or different

758.
dialog styles, when going from one function to the other. For
instance, data manipulation is based on the same schema
representation as the one used at schema deftnition time [5, 7, II,
13, 14, 30].

Database and Expert Systems Applications.pdf

F02 has been developped using a bootstrap approach too and we

759.
have defined a FA for every key concept of our model. The Schema
and Rule FAs are only used for the implementation of F02 and an
application programmer cannot use them to write FA_expressions.
This is justified by the re quirement for safe manipulation of the data.

Database and Expert Systems Applications.pdf

760. Given the formal query language described in the previous section,
we can now formulate the earlier queries as follows: "I want all
documents on robotic vision systems." match( document( aILnodes),
robotic I vision I system» "I want more information like this." top( 10,
subtract( rank( match( alLnodes, this_data), this_data), historylist»
 457 "Show me a document about vision that is related to one that I
have already looked at." top(l, document( match( relatecLdocs(
historylist), vision) » "Show me a document about vision that is not
related to one that I have already looked at." subtract( document(
match( allJlodes, vision», related_docs( history list » "Show me the
node that I looked at before on robots." top( 1, match( historyIist,
robot» As well, we can simply browse: Follow a link.

Database and Expert Systems Applications.pdf

Keywords: CIM, scheduling, planning, constraints, con straint

761.
propagation 1 Introduction In this paper we present a system for
generating produc tion schedules for the German Processing and
Archiving Fa cility for ERS-l satellite data( D-PAF). The scheduling
pro cedure bases on constraint-guided scheduling, where one
represents the scheduling problem by a network of restric tions. This
representation allows the application of con straint satisfaction and
constraint propagation techniques for supporting the scheduler.

Database and Expert Systems Applications.pdf

2.1 Ontologies In philosophy an ontology is a particular theory about

762.
the nature of being or the kinds of existence. The following short
definition describes ontologies as used in our sce- nario. In the
understanding of this paper they consist of both schema and
instance data.

Lecture Notes in Artificial Intelligence.pdf

763. 1, a particular table type can be detected by analyzing its primary

key, foreign key(s), and sometimes the instance data. The detection
strategies are as follows:

OWL Ontology Extraction from Relational.pdf

764. 1 Introduction In the last few years several research efforts have
concentrated on describing databases as first order logical
languages. The expec tations of this research is to enhance the
capabilities of database management systems with the espressive
power of First Order Logic ( FOL) to provide powerful systems that
can be used in AI applications. In fact, while database management
 systems are able to manage efficiently large amounts of data, logic
affords both an appropriate representation scheme of the application
domain of knowledge and a computational model for intelligent
databases or knowledge bases { KB}.

Database and Expert Systems Applications.pdf

1 The 02 Model In object-oriented database systems. data are

765.
organized into objects. Each object has its own identity : data and
their behavior are encapsulated under this object identity.
Manipulations on data are done via methods attached to objects.

Database and Expert Systems Applications.pdf

766. 5 Summary In this paper we discussed the use of a knowledge-

based system for fault diagnosis in complex real-time engineer ing
devices. The characteristics of these devices require that the system
combine reasoning, database, and real time data acquisition
capabilities. Our approach to build ing such a system was to design a
generic system appli cable to a family of devices which could then be
custom ized for use with particular members of this family.

Database and Expert Systems Applications.pdf

Clearly, the two studies provided by Astrova [22] and Sequeda [29]
767.
represent the most relevant ones because they proposed many
require- ment that can act as best practices for building ontologies
from RDBs. On other hand, building an ontology based on an
analysis of relational data ( Migration of the instances) is addressed
in [21, 22, 28, 29].

A novel approach for learning ontology from Relational db.pdf

768. 2 Foundations 2.1 Ontology In the understanding of this paper an

ontology consists of both schema and instan- tiating data. An
ontology O is therefore defined through the following tuple: O : = (C,
HC , RC , HR, I, RI , A). Concepts C of the schema are arranged in a
subsump- tion hierarchy HC .

Lecture Notes in Artificial Intelligence.pdf

 In 02 [ Ucluse & all. 88]. users may access data within a
769.
programming language (C02. an extension of C) directly on the
complex values of objects. A complex object is an object with
attributes having for values other objects. eventually complex ones.

Database and Expert Systems Applications.pdf

The reason is that it is difficult for different users or even the same
770.
user at different times to describe the same thing identically be cause
they can use synonyms, generalize/specialize catego ries and so on.
Hence, the key to efficient retrieval is to automatically perform partial
or approximate match of the description of multimedia data to the
description of a user query whenever exact match is not possible. In
this paper, we propose an intelligent approach to approximate match
ing by integrating object-oriented and natural language un
derstanding techniques.

Database and Expert Systems Applications.pdf

6 Conclusion We have presented a method for conceptual modelling

771.
in a knowledge based environment. The approach makes use of the
ability of the SYL LOG Expert Database System to execute near-
English specifications directly over SQL databases. This capability
can help in bridging the gap between a conventional conceptual
semantic data model and the intended application.

Database and Expert Systems Applications.pdf

In order to design a system integrating, in a homogeneous way, the

772.
features of a DBMS and those of an application generator we propose
to extend the form notion [3], [8]. This approach breaks the
separation between application programs and the DBMS, and gives
the possibility to manage at the same time structural and semantic
aspects of data. The form notion allows designing such a system with
a top-down approach by taking the form notion at the interface level
and extending this notion in several ways.

Database and Expert Systems Applications.pdf

 6 Conclusion We have presented a data model that takes advantage
773.
of the simplicity of the relational model and has most of the fea tures
of the object-oriented data models. We developed the named value
concept which is more general than the concept of nested relation in
IQL [ AbKa 89) and value 02[ LeRi 89b). Meta variables are
introduced to implement information hid ing, data abstraction and
inheritance.

Database and Expert Systems Applications.pdf

The main topics of the paper were the diagnosis algorithm used and
774.
the representations for the knowl edge and the data required by the
algorithm. The struc tured data, such as the readings from the
sensors and the characteristics of the components of the device, are
kept in a relational DBMS. Tbe structural knowledge of the device is
represented in a tree-shaped semantic network and is used to guide
diagnosis process.

Database and Expert Systems Applications.pdf

775. The master module handles requests in the normalized form using
the command "get", which has the following syn tax get
<yy>/<mm>/<dd> <for> <src> <var> <hour> <lat> < Ion> <alt> 307
DB will use the parameters to locate the necessary file, initiate the
required subprocess and send it the requests through the pipe. The
slave returns the data through an other pipe under the form of a
character string. DB gives the final result to the user as a series of
records contained in a unique character string, where each record
has the form: I <value> <var> <time> <alt> <lat> < Ion> For instance,
the call get 91/05/01 0 efr tt 12 wmw wmw all normalizes: 1.19389 TT
301200 649 42.

Database and Expert Systems Applications.pdf

Also they may model different aspects of the same concepts. Thus
776.
integration of data has become an area of growing inter est in recent
years. During this process two problems arise: first, how can we
reconcile the differences between diverse local, and possibly
conflicting schema definitions (for example, name differ ences,
domain and type differences); second, how can we establish
relationships between two or more diverse entities in different in
 formation sources, that are semantically related when such kinds of
relationships are not expressed in their schema specifications.

Database and Expert Systems Applications.pdf

777. Since hypermedia information systems contain both structured and

unstructured data, the integration of data base and information
retrieval methods is an inevitable part of a multimedia retrieval
function. In addition, adequate index ing methods and a novel
similarity function are being developed. In order to study new
methods, a hypermedia information system with real-world data is
under development.

Database and Expert Systems Applications.pdf

In order to support our scenario we also need to represent task

778.
models that are sensitive to changes both in underly ing data sources
and in local data. We therefore have im plemented database change
monitors both within WS-IRIS and in the IRIS prototype. Primitives
have been added to OSQL to monitor changes of views of both local
and global data.

Database and Expert Systems Applications.pdf

The data from the H.LT. system shall be transferred into the INEKS
779.
database. A transfer program has to be developed which 1 H. 1. T.
Hannoversches Informationssystem fiir Tumordaten Hannover
information system for tumor data 2 MUMPS = Massachusetts General
Hospital Utility Multi Programming System / Wolters/ 567 converts
and transfers the H.I.T. data of the after-care centre into the
INEKS database.

Database and Expert Systems Applications.pdf

594 W.- T. Balke and K. Mainzer embodied knowledge either from

780.
long-term profiles, intention, situation, or domain, they have to be
considered on a lower level than the explicitly provided terms (i. e. as
soft constraints that may refine too large result sets, but can be
relaxed if empty result sets are retrieved). Recently [12] introduced a
system of integrating preferences in the form of strict partial orders
 with a simple “I like A better than B” semantics into data- base
queries.

Lecture Notes in Artificial Intelligence.pdf

781. Codes: X-Yes; Blank - No DESC -A coda that Indicates if the item
DGSC - A code that indicates if government Codes: Y-Yer; N-No
developed from policies and regulations, drive system rules while
system rules describe processing logic and data relationships to be
implemented by the application code. To facilitate the requirements-
driven approach, business rules and policies are generated from
external regulations and laws. Common practices and procedures
within the business enterprise are documented in the pIan dictionary
of the data model as shown in Figure 6.

A Requirement-Based Approach to Data Modeling and Re-

engineering.PDF

Conclusion To sum up, in this paper, we tried to gather the most

782.
important and contributing approaches in the subject of the mapping
of the relational database to ontology. We attempted to provide the
reader with concise overview of these approaches in terms of
identifying the main drawbacks that the researchers in this field are
faced as well as suggesting solutions. In addition, the biggest
contributions within this paper are the fol- lowing: (1) We propose a
new life cycle for ontology learning from RDBs based on the software
engineering requirements; (2) We describe a new method for building
ontol- ogy from Relational database based on the predefined life
cycle; (3) We add three new semantics that can be extracted from
RDB; (4) we suggest an evaluation process based on two categories
of metrics: (i) Conceptual Ontology.

A novel approach for learning ontology from Relational db.pdf

 e., transaction control, multi-user support, system administration
783.
support, etc.). Recently, there have been a few studies about non-
standard data model translations into the relational data model [L
Y87, B090]. In one case [L Y87], the analysis focused on a semantic
data model conversion.

Database and Expert Systems Applications.pdf

784. Since this model also include support for office document object
processing. we are trying to add features for storage of limited (in
nature) processing methods. Method signatures can be included in
present model. but we are seeking more sophisticated methods for
storing the operations that we want to perform on an object. The
underlined constraint is to build our data model on top of a
commercially available (nested) relational DBMS.

Database and Expert Systems Applications.pdf

In this paper we will assume that the data model underlying the
785.
individual information sources to be merged is a variant of the object-
oriented model which treats all parts of the design as objects thereby
reducing the complexity of the analy sis [6], [7J. This methodology
can also be applied to distributed heterogeneous information-bases if
some kind of object-oriented data model is chosen as the canonical
data model and schema homogenization has been attained [8J.

Database and Expert Systems Applications.pdf

3 Realization Aspects 3.1 The System Architecture The system

786.
architecture (figure 3) reflects the division of the access procedure as
shown in figure 1. First, at the access level, problems of
heterogeneity of hardware and database systems, as well as
problems of data distribution and location transparency have to be
solved. Up to the level of the com munication knowledgebases the
access process is supported by the KEElink software.

Database and Expert Systems Applications.pdf

787. 5 Active Task Models The DBMS architectures developed for

conventional databases are not always feasible for supporting active
 databases. In particular, disk based databases are often too slow for
ap plications requiring fast responses. Our architecture is a
combination of traditional persistent data representation for
conventional applications and main memory techniques for time
critical mediators.

Database and Expert Systems Applications.pdf

788. Also, the inheritance hierarchies produced are not compliant with the
mapping principles. From the proposed compara- tive analysis, the
study highlights that OntoBase outperforms DataMaster in the
creation of data, object properties, and hierarchy structures.
Although, like DataMaster, it does not produce any cardinality in
response to not null and null columns.

Ontology learning from relational database a review.pdf

789. Hence, by describ- ing the input /output of the subtasks of abductive
assembly we also spec- ify the knowledge required to use the
method. Simulation can be used to evaluate a hypothesis because
the simulation can reveal whether the hypothesis is possible given
the data about the device. Causal refine- ments of a category can be
deter- mined by simulating to de te rmine the possible outcomes of a
set of inputs to a device.

Task-Structure Analysis for KM.pdf

2.8 Integrity constraints The notion of integrity constraint has

790.
received con siderable attention in the relational framework but also
in semantic data models. What we want to do here is to show how
integrity constraints can take place in the Serendip framework based
on abstract data types and views.

Database and Expert Systems Applications.pdf

791. TA is well suited for data model Implementation The toolbox

approach is based on the analysis of the domain to implement for
extracting the objects that constitute it, with respect to their structure
and behavior. The theoretical basis on which any data model should
lie encompasses such a need for domain analysis. Data models
 constitute domains that are built "after" their components and axioms
are determined.

Database and Expert Systems Applications.pdf

792. The remainder of this paper is organized as follows. In section 2 we

describe the essentials of an object oriented data model, GOM (
Generic Object Model), which-as the name implies-generalizes the
most salient constructs of recently proposed [1, 15] models in one
coherent research vehicle. In subsequent sections the base model is
extended: in Section 4 subtyping and subtype substitutability is
included.

Database and Expert Systems Applications.pdf

The different steps, objects and methods necessary to fulfil this

793.
process of object mapping (dependent on object descriptions and
object related operations) must be defined in a KaVDB. There are
various approaches that can be applied to organize objects,
descriptive and operational data within a database (see Fig. 2).

Database and Expert Systems Applications.pdf

794. 1 Introduction Hypertext and hypermedia [ yHMD88, Hala88,

HaKa88, etc.] are based on a network or a graph data struc ture, as
are semantic networks [ Quil68, LeMy79] and some aspects of the
structural model [ Wied83]. Two advantages of graph models are
navigational access and their associative connection strategy.
Navigation has the potential to support users to get target informa
tion from ill-structured information spaces.

Database and Expert Systems Applications.pdf

795. Deductive DBS store knowledge only in unstructured bitstrings and

processed it in an internal form of rules by the query mechanism to
derive new data from stored ones. An external rule management that
stores knowledge in single parts and makes it available to users does
not take place. . However, such kind of knowledge management
would be beneficial for a coupling with an XPS: Besides the
 simplification of the application program this results in a higher
flexibility, because knowlegde managed by a DBS may.

Database and Expert Systems Applications.pdf

796. However, it is difftcult to formulate a method of automatically

completing media structure from the semantic structure. Then the
semtmtic structure that used processing data in the form of the func
relationship between the above-mentioned semantic structure and
the media structure is formed, and an analogical matching is
formulated again, as shown in Fig. 2. The func relationship in this
ftgure describes a method of media processing.

Database and Expert Systems Applications.pdf

Proceedings of the 20th VLDB Conference Santiago, Chile, 1994

797.
Christian T. Ramiller Hughes Information Technology Corporation
MS SClS64lC41O PO Box 929 19 Los Angeles, CA 90009 1.
Introduction Re-engineering is the process of analyzing, upgrading,
and integrating enterprise information systems to meet the expanded
operating requirements of the present, as well as to prepare a sound
base for effectively meeting future needs. In this paper, we report our
experience, key technical approach, and lessons learned from
applying our framework ([ MH94], [ AHR94], [ AHR93], [HI93a], [HI93b])
to eight large scale legacy information systems, each containing
several million lines of code, and thousands of data elements.

A Requirement-Based Approach to Data Modeling and Re-

engineering.PDF

Additional rules are also captured in the plan dictionary through

798.
analysis of the application code (these are the system rules - that
implement the business rules). Rules are then linked to the model
views to which they apply as illustrated in Figure 5. Physical data is
correlated with business processes (model views) via the create,
read, update, delete ( CRUD) matrix in spreadsheets.

A Requirement-Based Approach to Data Modeling and Re-

engineering.PDF
 Here, the process of mapping each of tuple set as an instance of the
799.
corresponding OWL ontology class of the table was just the
generating process of an OWL ontology individual axiom, the concrete
was: map the corresponding tuple set value of non-foreign key
column of each tuple set as the value of the corresponding data type
property of ontology individual, so as to describe the relationship
between the two individuals. Among them, individual identifier was
the OWL class name that this individual belonged to_ the
corresponding tuple set value of primary key of the tuple set, if this
table had m columns, then the individual axiom generated by each
tuple set was m.

An Improved Ontology Learning Algorithm from Relational Schema.pdf

800. Here, the process of mapping each of tuple set as an instance of the
corresponding OWL ontology class of the table was just the
generating process of an OWL ontology individual axiom, the concrete
was: map the corresponding tuple set value of non-foreign key
column of each tuple set as the value of the corresponding data type
property of ontology individual, so as to describe the relationship
between the two individuals. Among them, individual identifier was
the OWL class name that this individual belonged to_ the
corresponding tuple set value of primary key of the tuple set, if this
table has m columns, then the individual axiom generated by each
tuple set was m.

An Improved Ontology Learning Algorithm from Relational Schema.pdf

801. We believe that our system provides a simple and elegant approach
to both retrieval of multimedia data and query specification. The
simplicity of our retrieval method lies in exploiting the semantics of
generalization and specializa tion abstraction of the object-oriented
model; the simplicity of the user interface lies in the natural way of
query specifi cation being directly obtained from queries expressed in
natural language.

Database and Expert Systems Applications.pdf

802. Enforcing Consistency through Encapsulation In the ERC+ data

model, the consistency preservation means enforcing denomination
 axioms and the structural invariants between objects. FOT example,
adding an attribute to an entity also registers ’the attribute as
belonging to the entity's schema. The consistency is enforced using
the C++ constructor mechanism which allows for the control and
enforcement of the initialization.

Database and Expert Systems Applications.pdf

803. VISUAL DATABASE and VISUAL USER INTERFACE In the preceding

section it has been stated that usually there is a need for explicit
definitions in order to integrate and link visual objects with an existing
application frame consisting of conventional data types, such as text
and data tables. In most cases this process must be done in a
completely ’manual’ way. This means that there are only few special
cases where support of automatic methods can be used (e.

Database and Expert Systems Applications.pdf

1 Introduction Many different data models have been proposed to be

804.
used in conceptual design (see Hull and King [13:, Peckham and
Maryanski iI5;, and Tsichritzis and Lochovski [16: for ex tensive
surveys and tutorials), and are widely adopted in practical
applications. This has negative consequences on the possibility of
fruitful information exchange among dif ferent developers, especially
in large projects, carried out over long periods, as standards are
seldom fixed, and, when fixed, change with time. Also, analysts often
change job and company, and have to adapt to new models or
versions thereof.

Database and Expert Systems Applications.pdf

The data of the database is organized as a set of disjoint fragments

805.
(relations) based on the Overview of the Design Pro cess "The work
of Gunther Pemul ia supported by the Au .. trian Science Foundation
under the contract J0461 -PHY The universe of discourse is
represented by an ini tial global scheme R which consists of the rela
tionalschemes (representations of object types of real world), a set of
global dependencies to express refer ential integrity constraints, and
a set of user views representing the requirements of a set of users in
 the multiuser database environment. Thus R is defined by the triple: (
RS, GD, V) where RS = { RS1( AITRlo LD1), ... , RSn( AITR,.,

Database and Expert Systems Applications.pdf

806. The database management system of ISACS was required to be

flexible in its handling of data. Due to the heterogeneous types of
data which is included in the database, e. g. process related data,
data from operator support systems, man machine interface data
etc., and to ease the integration of the different parts of ISACS, the
use of a cormnercial system, with the necessary flexibility, seemed to
be the solution with the most favourable cost-benefit ratio.

Database and Expert Systems Applications.pdf

1 Preface The increasing use of innovative medical techniques and

807.
new diagnostic technologies results in a fast growth in the amount of,
above all, formated information and thereby to the necessity of using
computer-aided information technology as data base management
systems.

Database and Expert Systems Applications.pdf

808. The SAGE system is essentially a global-schema multi-database

system [5]. It combines the unified database by masking multiple
databases as if they were one. The main interfaces developed on the
query engine use text fields to search the proc- essed data.

Lecture Notes in Artificial Intelligence.pdf

809. Fig. 2. Functional components involved in the pathway management

lifecycle Workflow control, pathway monitoring and documentation of
actions and devia- tions should not be distributed among different
applications because these function are closely linked together and
would need a complex and high frequent information exchange. At
best these functions are implemented in the HIS where knowledge
about treatments is available, where actions can be connected with
HIS functions like refer- ral, electronic order entry and time
scheduling and so multiple data entry can be avoided.
 Lecture Notes in Artificial Intelligence.pdf

XIMSA’ architecture is based on 4 distinctive modules (module

810.
interactions are presented in figure 1): − the XIMSA database stores
symptom and drug data related to self-medication, − the XIMSA
ontology uses the OWL [8] syntax to represent terminological (usually
called the Tbox) and assertional (called the Abox) knowledge in a
self- medication context., − the Simplified Electronic Health Record (
SEHR) stores information (in an XML syntax) concerning data such as
clinical antecedents and the history of drug consumption for a
particular patient, − the inference engine makes deduction with
respect to the XIMSA ontology, the patient's SEHR and the data
acquired during the navigation within XIMSA [6].

Lecture Notes in Artificial Intelligence.pdf

811. The whole process of modeling the workflow and dataflow is done in
a graphical user interface in the Lixto Transformation Server.
Graphical objects symbolize components, such as an integrator for
the aggregation of data or a deliverer for the transmission of
information to other software applications. By drawing connecting
arrows between these objects, the flow of data and the workflow are
graphically defined.

Lecture Notes in Artificial Intelligence.pdf

Conventional information systems use relational databases for data

812.
storage and management. Relational databases require a stable data
model. In some cases, knowledge of the system is dynamic and it is
necessary to deal with its evolution.

Formal Ontologies in Information Systems.pdf

INTRODUCTION In recent years there has been considerable interest

813.
in building deductive databases, which support intelligent inference
as well as straightforward retrieval of data. These systems have
almost ex clusively used the relational data model as a starting point
[6, 8, 14], which is natural in view of that model's theoretical
foundation on first order predicate logic. Most of the research in this
 area has focused on the optimization of queries over recursively-
defined rela tions [1-5,7, 8, 12, 13, 16].

Database and Expert Systems Applications.pdf

814. The operation for an object is executed by sending a message to it.

When the object receives a mes sage, the method corresponding to
the message is invoked and the object behaves like the modeled
entity. A class is like a template such that similar objects can share
method implementation codes and data structure.

Database and Expert Systems Applications.pdf

The following sections will explain concepts and techniques

815.
applicable to the domain of explicit visual object representation.
Section 2 will motivate the use of databases for a common
management of visual objects. Section 3 shows general
n:quirements and consequences for the integration of visual data
objects in different application environments.

Database and Expert Systems Applications.pdf

However, current hypertext systems have several well-recognized

816.
problems[2], especially user disorientation in a hypertext network and
the cognitive overhead for creating, managing and choosing links[l].
Much research has been devoted to solve these problems but ouly a
small part of this has been concerned with using database
technology. DBMSs have powerful abilities for managing structural
data[3].

Database and Expert Systems Applications.pdf

817. We use the mediator concept to extend current DBMS tech nology to
support such distributed, heterogeneous, and dy namic
environments. Mediators make it easy to ’plug in’ new data resources
once there is a public interface protocol. In this work we focus on
active mediators, where the appli cation instructs a mediator to
actively monitor databases for changes to information that the
application depends on, and provide primitives for applications to
adapt to these changes.
 Database and Expert Systems Applications.pdf

However, the clinical relevance and utility of these findings await the
818.
results of prospective studies. We see our main contribution in this
paper in introducing and applying a many-sided analysis approach to
real-world data. The application of diversified DM techniques, which
are not necessarily accurate and do not best suit to the present
problem in the usual sense, still offers a possibility to analyze and
understand the problem from different perspectives.

Lecture Notes in Artificial Intelligence.pdf

In Figure 1, an example of structure specification of the entity

819.
CLIENT, identified in a credit management application, is presented.
Attributes in Figure 1 are the ones necessary to charac terize a client
in terms of credit risk and to know his state of orders and debt at any
given moment, where the specification of data types ORDERS and
INVOICES are: ORDER_ DETAIL= recordof(product_code : oid
PRODUCTS; quantity : integer <= 999);); ORDERS = recordof
(ordecnumber : integer <= 9999; ordecdate : DATE; detail : setof(
ORDER_ DETAIL);); INSTALLMENTS= recordof(installment_no:
integer <= 99; amount : integer. <= 9999; due_date : DATE;
installment_status: integer in [1,3];); INVOICES =
recordof(invoice_number: integer <= 9999; invoice_date : DATE;
inv_detail : setof ( INSTALL MENTS);); Notice that invoices are paid in
installments, each with a given due date, and that ORDERS make
reference to PRODUCTS, which is another class with the obvious
attributes of producccode, description and price.

Database and Expert Systems Applications.pdf

820. Behavior Specification As stated before, we attempt a behavior

specification which is as non-procedural as possible. In doing this we
use logical-based preconditions, whenever is possible, to indicate
processing rules that should be satisfied previous to the execution of
actions associated to a method. Hence the general structure of a
method is the one shown in Figure 2, where "->" indicates that the
method is applied to a given object of a class; and, alternatively, "=>"
that the method is applied over a set of objects in a class;
parameters of the method can be recordof, setof or objects; there is
an optional data type returned by the method;under local there are
 variables that participate in the method and their corresponding
types; and preconditions are.

Database and Expert Systems Applications.pdf

821. The so called yinual yisUIJI objects resulting from this process are
organized in a knowledge augmented yisUIJI database system. In
this way visual data can consistently be accessed and used by
different applications, without primary impact of physical constraints
given by a specific application environment KEYWORDS Visual
databases, visual object description, visual object representation,
visual knowledge representation 1. INTRODUCTION Manifold
progress in the domains of knowledge engineering, of hardware and
software techniques is a challenging basis for the development of
increasingly complex application systems.

Database and Expert Systems Applications.pdf

( ) mmRelTab T T is a many-to-many binary relationship table ( )

822.
omRelTab T T is a one-to-many binary relationship table ( ) ooRelTab
T T is a one-to-one binary relationship table Supertype/subtype
relationship ( , ) subOf T R A subClassOf relationship between a
subtype entity table T and a supertype entity table R Keys and data (
) NonFkSet T The non-foreign key attribute set of a given table T ( )
Data A The values of given attribute(s) A

OWL Ontology Extraction from Relational.pdf

Abstract Currently, semantic integration became an attractive area in

823.
several disciplines, such as information integration, databases and
ontologies. As many useful data are stored in the existing database
management system of enterprises, the ontology learning method
could be used to convert relational database to ontology. Therefore,
one of the main challenges in the research of data integration and
sharing based on semantics is to construct the mapping between
relational databases and anthologies.

An Improved Ontology Learning Algorithm from Relational Schema.pdf

2 Information management in the mining industry The main tasks of

824.
In< ITkscheiders are to provide sufficient know ledge about the
 location and quality of a deposit, to determine, whether it can be
exploited technically resp. economically, and finally to plan the
working and provide the basis for safe and undisturbed operation of a
mine (cf. [1]): - evaluation of a deposit - planning and designing a
mine - surveys of seams and all structures underground -
representation of all quality parameters, geological, anti geometric
infonnation about the deposit and the mine in maps and manuals -
calculation of expected coal output - recognition and prevention of
risks - assessing dislocations and damages caused by the working In
order to evaluate a deposit a lot of geometric, physical, geological
and chemical. infonnation is needed. The quality of a deposit can be
described by the following parameters, which are determined by
samples from special drillings as well as from regular production,
analyzed in laboratories: - quantity and thickness of coal seams -
ratio of seams and barren-ground - rate of sulphur, cinders, and
volatile components of the coal In addition to the material properties
the geological constellation of the deposit has to be inspected: depth,
thickness and dip of the different .

Database and Expert Systems Applications.pdf

If class C2 is a subclass of class C1, then the data structure and

825.
method implementation codes are inherited to each instance object
belonging to C2.

Database and Expert Systems Applications.pdf

So far, we represented any local similarity measure for a symbolic

826.
data type as a similarity table and conducted the optimisation on
corresponding matrices (consisting of φ2 entries for an attribute A
using a data type with | DA| = φ elements in its domain). However,
local similarity measures for attributes based on taxonomic and
ordered symbolic data types are of course definable with respect to
the characteristics of those types: This property will be exploited by
the concepts we introduce in the following.

Lecture Notes in Artificial Intelligence.pdf

827. Storing objects in our sense in Starburst leads to another problem: at

the level of the cluster index a record may completely fill a datapage
or exceed the page capacity if there are some long attributes with
 descriptors of maximum length in additon to the other attributes.
EXODUS seems to be suitable for the requirement of individual
access to any portion of data. However EXODUS is guided by the
idea that a long object covers a relation structure (or something
similar). which is similarly handled in WiSS.

Database and Expert Systems Applications.pdf

828. IS Development Life Cycle Information Technology (IT) Task Use of

Requirement-Based Re-Engineering Products I Business
Requirements Information system Requirements Software/ Database
Requirements Preliminary Design Detail Design Implementation Unit
Test Integration * Functional Economic Analysis l Business Process
Improvement l Enterprise Data Architecture * Enterprise Data Model l
Cross-functional Model Integration l Enterprise Model Integration l
Data Element Standardization l Enterprise Data Warehouse l
Enterprise Data . Repository l Enterprise IT Consolidation Planning l
Enterprise IS Security l Forward Engineering l Object Engineering l
Data Quality Engineering l Database Generation l Screen Generation
l Report Generation l Data Migration l Data Quality Assurance l Data
Administration l Training l data requirements specified in the re-
engineered data model l understanding of requirements and how tbe
IS implements the requirements I l Maintenance I Table 1. Use of
Requirement-Based Re-Engineering Products new application
architecture and technology l modifying one application causes
errors, aborts, and / infrastructure. or erroneous information in a
different application As a part of our re-engineering approach, we
resolve data A requirement-based data modeling and re-engineering
conflicts (e.

A Requirement-Based Approach to Data Modeling and Re-

engineering.PDF

I~ t~e p~ocess of designing and implementing applIcatIOns m the

829.
field of resource allocation (school timeta,hles, software project
management [l], crew formatIOn for bus driving [4]) we have
discovered that the ability of performing relational operations on the
data is by no means less important than performing computations on
the data in a rule based style.
. The l. iterature offers a few papers which present mnovahve
spreadsheets and share some of the motivations of our work,
although all of them stress the programming techniques of the
 spreadsheets rather than the manipulation of the spreadsheet itself.
The Logicalc system [10] is implemented in Prolog and offers
deductive ~eatures derived from it.

Database and Expert Systems Applications.pdf

e., classes, properties, axioms). During the translation process, they

830.
also provided user assistance when some kind of ambiguities occurs
and domain semantics cannot be inferred. In order to discover more
hidden semantics embedded within a RDB than that of [9], Astrova
[10] proposed an approach to reverse engineering of RDB to ontology
by an analysis of keys, data, attributes correlations as well as their
combination.

OWL Ontology Extraction from Relational.pdf

The purpose of the role model was to contrast this approach with
831.
conventional record-based database systems where logical records
were used to represent all aspects of modeled entities and to break-
out the one-to-many relationship pattern, carried over from
hierarchical and network data models. The role data model
introduced a static part for modeled ob jects called the entity, which
was derived from its corresponding entity-type, and a dynamic type
called the role-type. An entity established existence, while the role
type established behavior for that entity.

Database and Expert Systems Applications.pdf

Research into the causes of all kinds of cancer disease is at present

832.
of a largely speculative nature and, applying conventional
documentation techniques, gives no reason to expect any greater
success. It is therefore essential to store in computer-assisted cancer
registers extensive data about the real disease frequency in the
single areas and to use modern information technology (database
systems, expert systems) for analysis. Also, often discussed
environmental factors (different forms of nutrition, consumption of
alcohol, abuse of nicotine, radiation influences, especially by nuclear
power stations) can only be recognized if more exact data about the
disease incidents in a region are recorded and can be analyzed
together with the supposed pathogenetical factors.
 Database and Expert Systems Applications.pdf

1.2 Problem Definition Powerful tools for Extracting, Transforming

833.
and Loading ( ETL-tools) data from source systems into a data
warehouse are available today. They support the data extraction from
internal applications in an efficient way. But there is also a growing
need to integrate external data, such as market information, into
these systems.

Lecture Notes in Artificial Intelligence.pdf

(2021), which are the newest, are those which try to propose a more
834.
detailed translation between SQL and OWL patterns. However, from
our analysis can be seen that each method presents some
shortcomings in its ontology generation and that, after about 20 years
of work in the field of Ontology learning starting from RDBs, no con-
vergence has yet been found on the best way to convert the data
source into a valid ontology. For this reason, a merging of the various
rules and advantages of the approaches can be desirable in order to
be able to carry out an automatic generation of an ontology that is as
useful and faithful as possible to the semantics implicit in the original
data source.

Ontology learning from relational database a review.pdf

835. In order to achieve a content based retrieval, in our approach, we

use natural language captions which allow the user to describe the
contents of multimedia data. In a similar manner, users will specify
their queries on multimedia data contents in natural language form. A
problem is that different or even the same user describe the same
thing differently at differ ent times which results in the descriptions of
the contents of mul timedia data to rarely exactly match the
descriptions of the user queries.

Database and Expert Systems Applications.pdf

As mentioned in the beginning, we are mainly interested in the se

836.
mantics of environmental datil For this reason, we introduce the
object representation level in our model: Here the seman tics of the
objects stored in the database is represented. For ex ample, at this
 level, a series of measurement results should be represented as a
pointwise specified function, and the spatial and temporal distribution
of these values also should be repre sented, e. g. for allowing the
interpolation of values.

Database and Expert Systems Applications.pdf

To extend functionality of expert systems with regard to ac cess to

837.
different autonomously managed data an interface for logical data
integration is developed (fig. 3). The definition of application
dependent schemes and user defined operations abstracts from
original data representation. Goal is the transparent use of data and
operations from e.

Database and Expert Systems Applications.pdf

In doing so the modeling follows the general paradigm of "object

838.
orientation". While this could in principle also be done within the E/R
model, extension concepts, such as object identity, generalization /
specialization, aggregation on object level, multi valued attributes,
grouping, primary key constraint, referential integrity constraint, allow
a much closer orientation at a structurally object oriented
representation of an application domain. The selec tion of ExER also
in principal, allows the later integration of databases founded on
other data models ([4, 5]).

Database and Expert Systems Applications.pdf

330 2.2 Top down and manual The development of the central
839.
representation scheme based on the ExER model offers the
possibility to express explic itly, information about data and
relationships included in data - even across different databases - on
a semantically high level. Otherwise, this knowledge is in many
cases di rectly coded in applications. New applications on distributed
heterogeneous knowledge sources can be realized within the chosen
approach, preserving the logical independence from data in a better
way.

Database and Expert Systems Applications.pdf

 Additionally, it corrects and adds the contents of a list form already
840.
stored into the domain D2. The media structure changed in this step,
and the relationship between instance defmitions, ako, is_a and
part_of and so on, is the conversion of data registered to OMS into a
list form. The func relationship is composed of ref function to give it
an offset that is applicable to each data.

Database and Expert Systems Applications.pdf

ADAMS: an Aggregate Data Management System with Multip

841.
Interaction Techniques Fernando FERRI*, Patrizia GRIFONI",
Leonardo MEO-EVOLl", Fabrizio L. RICCI" " Consiglio Nazionale
delle Ricerche, Istituo di Studi sulla Ricerca e Documentazione
Scientifica; Via Cesare de Lollis 12; 00185- Roma; Italy * Universitit
di Roma " La Sapienza", Dipartimento di Informatica e Sistemistica;
Via Salaria 113; 00198- Roma; Italy ABSTRACf The paper describes
ADAMS ( Aggregate Data Management System), a system for the
definition and manipulation of Statistical Databases ( SDB). ADAMS
adopts different interface approaches according to the user profile.
The System displays data by means of spreadsheet and simplify the
user result analy sis activity.

Database and Expert Systems Applications.pdf

Palazzolo N., - Tammaro A.M., Studio di un sistema in tegrato per il

842.
recupero dell'informazione bibliografica ne! campo dei diritti
dell'Antichita, in Proceedings of IV In ternational Congress on the
subject " Informatica e regolamentazioni giuridiche", II, 28. ., T.C.
Tan, M. Smith, M. Pegman, Conceptual retneval usmg object -
oriented approach, Proceedings of DEXA90, 1990 Petrucci P., The
Thesaurus as a relational data model In Automated analysis of legal
texts, ed. by Martino A., Socci F., Amsterdam North Holland, 1986. L.
Kerschberg (ed.),

Database and Expert Systems Applications.pdf

Fig. 3. The Microsoft PowerPoint document created by the test

843.
subject in response to a request to produce a short (five-minute)
presentation to corporate executives on the impact of RFID
technology on supply chain management defined the task (based on
folders created in TaskExplorer), gathered information from the Web
 (based on the Internet Explorer history), and used that information
(based on TaskTracer’s data on clipboard and application usage).
The next step is to analyze the data to reveal the knowledge-
intensive processes of interest.

Lecture Notes in Artificial Intelligence.pdf

The new knowledge management solution was integrated in the

844.
existing infrastruc- ture. For this purpose, a database of studies and
patents as well as other project data- bases with specific values from
other investigation series were integrated in the mod- ule “library” (
DMS). The semantic network, which is automatically generated from
this information, provides a simple management of information as
well as the recogni- tion of relations between heterogeneous topic
complexes.

Lecture Notes in Artificial Intelligence.pdf

N is a finite name set partitioned into: (1) a subset ET of entity table

845.
names; each entity table contains rows of instance data describing
entities in the real world, (2) a subset RT of relationship table names;
each relationship table contains rows of instance data describing the
relationships between entities, and (3) a subset DT of datatype
names; each datatype is a predefined RDBMS datatype, specifying a
value range of the relevant instance data.

OWL Ontology Extraction from Relational.pdf

The technical barrier, which is caused by the lack of interfaces

846.
between a traditional language like FORTRAN and modern tools for
knowledge based systems, may be overcome in this way, because
interfaces to standard data manipulation languages (esp. SQL as
quasi standard) are rather common. To give an example: the expert
system for risk detection in mine planning, which is currently
developed using a high level KBS tool (c.

Database and Expert Systems Applications.pdf

847. Meta Variables and Inheritance in an Object-Oriented Data Model

Yanjun Lou and Z. Meral Ozsoyoglu Department of Computer
 Engineering and Science Case Western Reserve University
Cleveland, OR 44106, USA Abstract In this paper, we discuss the
properties of an object oriented data model that combines the
simplicity of relational data model and features of object-oriented
data models. Meta variables are introduced for information hiding,
data abstrac tion, and type inheritance. Type inheritance hierarchies
are built up by instantiating meta variables in types.

Database and Expert Systems Applications.pdf

2 Office Automation Database management in office automation

848.
concerns the captur ing, archiving, and dissemination of documents
for decision making purposes [9], [21]. These document databases
are large collections of semi-structured objects, that range from
simple bit maps, ascii strings, and letters, to highly regular data entry
forms and EDI messages. Furthermore, collections of documents are
structured in two (independent) ways: a hypertext structure and
folders.

Database and Expert Systems Applications.pdf

When a relationship type is browsed, the corresponding data base

849.
system calls will be invoked to check what objects are . related with
each other. Such information will be displayed on the screen for the
user to browse. We would also like to define InterSect _ DM using a
formal notation.

Database and Expert Systems Applications.pdf

850. In our opinion, the data type defined in this way, which from now on
will be called spreadview, combines the advantages and the features
of the relational and of the spreadsheet representation/programming
models.

Database and Expert Systems Applications.pdf

2.3 Review In the following, we describe the papers subject of this

851.
review. For each article, we will briefly indicate the type of input and
output of the method proposed in the papers, according to the
previously described classification: Li et al. (2005) propose an
 approach for learning OWL ontology from data in relational databases
using a set of mapping rules.

Ontology learning from relational database a review.pdf

852. Table 4: field table ATTRIBUTE meaning field lD identifier of field field
name field name(,- attribute name in DB) data type for example char,
text, date and so on.

Database and Expert Systems Applications.pdf

853. Several approaches based either on extended relational models or

on new, object oriented techniques are currently under development
(cf. for example: [11], [12], [13], [14], [15]). They especially offer the
possibility to use - or even to define one's own - geometric data types
and support queries including geometric operations like "inside" or
"intersects".

Database and Expert Systems Applications.pdf

854. The concept of type in our model is used to model the structure part
of data. Every class name is associated with a type. This type is
called the ,chema of the class.

Database and Expert Systems Applications.pdf

In 02, values as well as objects can be used to construct more

855.
complex objects. IQL [ AbKa 89] introduces relations into its data
model, 80 that relational data model is prop erly included. Extending
this idea, we introduce ’named val ues to further address the
importance of values in object oriented data model.

Database and Expert Systems Applications.pdf

448 seek directive. When LEONARDO executes a knowledge base it

856.
literally seeks for a value for the object named in the seek rule. In
backward chaining it is assumed no data and only ask for the
information needed to prove the evidence of a line of reasoning,
 therefore the system in turn receives an answer from the user and
returns to evaluating a rule.

Database and Expert Systems Applications.pdf

857. 3.3 Prototype for experimentation The main clinical purpose of this
system is to help optimise the treatment of patients with peripheral
vascular disease. Our aim is to improve on the support offered in a
conventional data base system by providing flexible assimilation and
retrieval facilities to help clinicians involved in the decision making
process. We are investigating various strategies for applying the
knowledge in the system, including temporal knowledge, for guiding
retrieval in a helpful way.

Database and Expert Systems Applications.pdf

Each of them is reviewed from the point of view of their compliance

858.
with normal forms to assess the possibility of using a particular
method when developing databases for storing graphs. It is shown
that for such a task, each of these methods has a large number of
disadvantages that lead to low efficiency of both data storing and
processing. The article suggests the way to represent graphs in the
form of a relational list of edges corresponding to the third normal
form and allowing to eliminate the disadvantages of other methods.

Graph to RDBMS.pdf

This representation accepts any combination of "fact" but makes it

859.
difficult to verify data integrity and to process a large volume of data
[13]. In addition, an object property is often converted to a join table,
and a data property is always converted to an attribute. These
conversions rarely take into account the value of the cardinality.

Generating_Relational_Database.pdf

860. Type of Knowledge 1o Organization best practices for coding 2o

Organization best practices for tests 3o Organization best practices
for software documentation 4o Organization best practices for quality
assurance 5o Knowledge about the organization software processes
5o Organization best practices for the problem solving process 7o
Organization best practices for specification and analysis of software
 requirements 8o Organization best practices for software. design 9o
Knowledge about the application domain 10o Organization best
practices for project management 11o Organization best practices for
system engineering 12o Templates of documents, including real
examples of their use 13o Organization best practices for the
evaluation and improvement of software processes 14o Software
items such as specifications, system architecture, source code and
test cases 15o Organization best practices for process modelling.
16o Organization best practices for software maintenance 17o
Knowledge about the evaluation of the organization software
processes 18o Knowledge about the distribution of competences
among the organization professionals 19o Reports on the best
practices in the software industry 20o Knowledge about
organizational objectives and goals 20o Organization best practices
for training 22o Organization historical data related to project
management 23o Organization.

Lecture Notes in Artificial Intelligence.pdf

For example using property inheritance in an expert database sy

861.
stem is just one special reasoning mechanism and others are
possible and might be demanded by the users in the future. By
contrast, SYLLOG offers the user an easy way to specify any
reasoning mecha Illsms or semantic, data constructs they find
necessary. The result is an open system, and together with the
modular technique descri bed lets the user construct tailored
information systems.

Database and Expert Systems Applications.pdf

 Although a number of data- base systems exist on the World-Wide-
862.
Web, which claim to help you identify experts that match a specific
profile, most of these tools rely on people to self-assess their skill
against a predefined taxonomy. On the other hand, while a number of
search engines are available on the web, the entity seeking for an
expert has to use a combination of different tools in order to find the
appropriate information. With SAGE, all the infor- mation is easily
accessible due to the versatility of its searching options, which allow
users to refine the search until they obtain the degree of accuracy
required.

Lecture Notes in Artificial Intelligence.pdf

KnowBiT – Knowledge Management in the Biotechnology Industry

863.
725 which information objects should be integrated in the new
solution, was made to- gether with the users. It was then decided that
the project databases will not yet be integrated due to the sensitivity
and structure of the data contained. All other docu- ments (a total of
more than 100 GB) were taken over into the document management
system windream, so that the work done so far was adapted with
group drives.

Lecture Notes in Artificial Intelligence.pdf

object description b · / . o Jee! representatIOn ,b j,l, ’:::::""’0

864.
environmental data I environment -logiclll query statement /~""""’; OO
- pres. specificatiot qry environmental problem Figure 1: Conceptual
model Figure 1 shows our conceptual model for EIS. Observations
~; the environment produce environmental data, which can be
measurement results or data derived from these values, IIA well as
written reporh or even multimedia documents with graphics, photos,
voice annotations or videos. Current work on multimedia information
systems focuses on the storage of multimedia objects [46] [24] and
on transformations of the con ceptual and the layout structure of
multimedia documents.

Database and Expert Systems Applications.pdf

865. 3.3 Knowledge Sifter Meta-model The previous sections have

described how the cooperative agents and web services support the
search for relevant knowledge from both local and open-source data
 sources. The end-to-end scenario shows how the various agents and
sources interact. The OWL schema is instantiated with information
regarding query and its various transformations into the final ranked
results.

Lecture Notes in Artificial Intelligence.pdf

For each T ET RT∈ ∪ , there is exactly one primary key ( PK) ( )pk T
866.
whose values uniquely determine each row of the instance data in T ,
where either ( ) ( )pk T attr T∈ (in this case ( )pk T is a single-
attribute key) or ( ) ( )pk T attr T⊆ (in this case ( )pk T is a composite
key with more than one attribute).

OWL Ontology Extraction from Relational.pdf

The ALDOUS90 Project : Merging Object-Oriented Databases and

867.
Knowledge-based Systems F.- Y. VILLEMIN, A. PAOLI, I.
TOURRILBES & M. LE CENTRE D'ETUDE ET DE RECHERCHE EN
INFORMATIQUE DU CNAM 292, rue Saint Martin 75141 PARIS
CEDEX03 ( France) ABSTRACT The ALDOUS! IO project Is originated
in the dimcultles met in realizing complex expert systems (in CAD,
CASE, ) using commercial shells such as KKE : they provide a good
set of faclUties for devel9ping experl systems, but as the amount of
data increases, their performance drastlely decreases. Object-
oriented databases, such as 02 of the GIP-ALTAIR, optimize object
management which Is not done in these shells.

Database and Expert Systems Applications.pdf

868. The Underlyin& Relational Data Model As mentioned in one of the

earlier chapters, we have used a standard RDBMS to store the objects
of C+ + classes into the database. Therefore we have had to take
care of transforming the class definitions into a data model consisting
of flat relations only. This is trivial for simple classes containing only
attributes with standard basic data types.

Database and Expert Systems Applications.pdf

869. e. if the argument term is of the form f (t" ... ,til where f is the name of
a data or object function then !’( S!.., il) is defined as follows: 2.4
 Query Terms The following list may give an impression what kinds of
queries on database behaviour are desirable.

Database and Expert Systems Applications.pdf

870. In the second step, transformations to the export schemas

homogenize the component schemas by resolving schema conflicts.
The categories of potential conflicts depend on the common data
model; a typical example is a concept that is represented by an
attribute in schema A and by an object in schema B. In the third step,
the export schemas are subject to a merge operation that builds on
semantic relationships between concepts of the different schemas.
This yields the federated schema which is still expressed in the
common data model.

Lecture Notes in Artificial Intelligence.pdf

871. 3) Illustrating more semantic information to receive theoretically by

making relational analysis to primary key, data and attribute, and
completing the transition to logic through making use of certain of
mapping rule.

An Improved Ontology Learning Algorithm from Relational Schema.pdf

was a set of data-type property identifier; DT was the set of data-type

872.
identifier (the identifiers of each data type was that of the predefined
XML schema used in OWL ontology); A R was the set of an object
property identifier; I was the set of an individual identifier; X was an
axiom set; What needed to point out were: the complete syntax
format of all identifiers was a URI reference, which was consisted of
an absolute URI or prefix expressing namespace, and a fragment
identifier. (the paper omitted the part of namespace for easy). Each
axiom can be realized by a number of constructors acted on identifier
or description.

An Improved Ontology Learning Algorithm from Relational Schema.pdf

Introduction VLSI circuits are becoming more and more complex,

873.
and good CAD software tools are essential in their design. A database
 management system can playa central role in storing and integrating
design data, enabling a quick development of new CAD tools.

Database and Expert Systems Applications.pdf

874. 3 Previous support by computer aided methods Having summarized

the complex tasks of mining engineers and especially of the
markscheiders let us now turn to the previous computer support for
these tasks for fixing clearly the starting point of the information
system described in this article. In order to support and improve the
manual processing of deposit data since the 1970’s computer
programs were developed for several specific mining applications [3].
(Most of numerical calcula tions and CAD-programs are implemented
in FORTRAN.

Database and Expert Systems Applications.pdf

875. While the first question is only relevant for the users of the specific
application, the second one is directed towards the development of
future EIS. By collecting enough evidence for the necessity of certain
concepts, the research activity in this field can be stimulated. In a
related application area, we have performed empiriclIl st udies in
order to identify relevant concepts and to propose ’the design of
future information systems: In the project " Ac cess to materials data
banks: user studies and system de sighn", we have regarded
materials data banks ( conlRining values of properties of materials)
[I].

Database and Expert Systems Applications.pdf

 Introduction. As the utilization of information systems grew. the need
876.
arose for adapting them on one side to the new kinds of applications.
and on the other side to the ever growing user population. These new
necessities made the information system.; evolve in several
interrelated direc tions:from the representation power standpoint.
new data models are being defined to extend the universe of real
world objects that can be handled by the information system. from
the usability standpoint. a growing interest is given to the problems
related to the human·com puter interaction.

Database and Expert Systems Applications.pdf

Let us discuss now whether our approach can be qualified as an

877.
object-oriented data model or not. We compare our model with the
main features and characteristics that a system must have to be
classified as an object-oriented database according to [2]. Our
system partially satisfies rule 1 (complex objects), rule 2 (object
identity), rule 3 (encapsulation), rule 4 (type and classes), rule 5
(class or type hierarchies), rule 6 (overriding, overloading and late
binding), rule 7 (computational completeness) and rule 9
(persistence).

Database and Expert Systems Applications.pdf

The DERINFO methodology, whose general principles with regard to

878.
juridical techniques are outlined above (a deeper treatment of which
would require developing juridical proposals which go be yond the
scope of this article), was drawn up by the Seminario de Informatica
y Derecho at the University of Zaragoza while developing a set of
programs, research and reports aimed at assisting the jurist to
access legal documentation, that is, to access the law. From a
computational viewpoint, one of the substantive bases of this work is
the prior existence of legal data banks. Before discussing what was
done, it should be pointed out that the characteristics of typical
democratic judicial prac tice (access, interpretation and application of
law; construction of dogmas) and the requirements of the legal
system of a democratic country, were res pected at all times [5] The
activities of the Seminario can be sum marised as follows: a)
Research leading to the construction of pro grams - some prototypes,
some working programs - to provide.
 Database and Expert Systems Applications.pdf

3 Assessing the methods of graph representation based on the

879.
normal forms The normal form is a relation feature in the relational
data model that characterizes it from the point of view of redundancy,
which can potentially lead to logically erroneous results of sampling
or changing data [6].

Graph to RDBMS.pdf

880. It has been challenging to convince management that re-engineering

is a substantially broader and more complex task than just
“restructuring the code” by pumping the code through a CASE tool.
Restructuring the code is useful to improve the system
maintainability, but it does not support the following critical tasks: (1)
facilitation of data migration, (2) evolution of a departmental system
to satisfy enterprise-wide system requirements, (3) data sharing
among functional areas (e. g., Health Care, Personnel, Pay), and (4)
incorporation of new or changed data requirements resulting from
business process improvement activities.

A Requirement-Based Approach to Data Modeling and Re-

engineering.PDF

g., synonyms, homonyms) during data approach should be used to

881.
re-document or map business modeling and model integration. In
section two, we requirements, functional requirements, and data
describe our approach to identify data conflicts and requirements to
architecture, design, and implementation. classify data conflicts.
Depending on the scope of the re- As depicted in Table 1, re-
engineered data models enable engineering effort, various levels of
integration will occur. identification of obsolete portions of
applications, We developed an integration taxonomy to guide our
outstanding unfulfilled requirements, applications that integration
process for data modeling and re-engineering require changes or
consolidation based on new functional legacy information systems.

A Requirement-Based Approach to Data Modeling and Re-

engineering.PDF
 The data structure of this class is similar to that of a fuzzy-object
882.
classes except that contents at tribute contains the corresponding
class name and the grade of the instance-of relationship. For users,
this FUZZY-IlI'STANCE-OF object and two objects (for Action and
Dramatic) are invisible, and a single Movie object belongs to these
classes with some certainties for users.

Database and Expert Systems Applications.pdf

Ontology study is also relevant to ELS in that they may be used to

883.
represent complex relationships between knowledge competencies
in a taxonomy. On the other hand, the third design concern has been
addressed through the application of artificial intel- ligence and
decision support techniques, for example text data mining. Web text
data mining could be instrumental in mitigating the need for ELS to
depend on expert’s self-assessment and taxonomy development, by
inferring from the employee’s Web pages their corresponding areas
of expertise.

Lecture Notes in Artificial Intelligence.pdf

Furthermore, a programme for the Stadtische Kliniken Oldenburg

884.
shall be developed to provide transference of data to the parent
documentation ’ Cancer’ of the National Health Department For the
support of epidemiologic cancer research, selected epidemiologic
retrieval classes must be worked out which are realizable with the
linguistic means of a database system or possible extensions. For
the manipulation of the data in INEKS, from the aspect of
epidemiological analysis, possible extensions of the data material
must be carried out, especially with regard to regional quantities of
influence. For all analy'sis results, graphical illustration should be
made possible in order to increase their transparency.

Database and Expert Systems Applications.pdf

885. 4) Connotative semantic information can be extracted through

making relational analysis to primary key, data and attribute, as well
as applying particular mapping rule, so the ontology can be obtained.
The level of
 An Improved Ontology Learning Algorithm from Relational Schema.pdf

A specific structure represents an instance of a real document.

886.
wheres a generic structure is a template of similar specific structures.
In [ MU89]. the ODA data model has been interpreted according to
principles of object-orientation. We have found that our approach
based on generalization and object sharing using nested relations
provides a suitable platform on which to build an object oriented ODA
data management system.

Database and Expert Systems Applications.pdf

4 Requirements for an information system It was demonstrated in the

887.
previous chapter, that the actual situation in mining information
management practise is characterized by a great variety of
heterogeneous, special purpose application programs, each using its
own data and a tailored data structuring. The resulting redundancy
and semantic overlapping of data stored in various file structures not
only cause problems regarding data availability and compatibility, but
also the risk of undetected inconsistencies between them. In order to
improve the technical support for markscheiders, it seems
indispensable to develop an integrated information system
coordinating all relevant data management tasks for the different
planning and assessment activities (cf. [8], [9], [3]).

Database and Expert Systems Applications.pdf

Semantic Ambiguity In the fuzzy relational data model, a fuzzy

888.
relation is defined as a fuzzy subset of Cartesian product of n
domains (each domain may be a fuzzy set). That is, each tuple in a
fuzzy relation is associated with the tuple's membership-function
value, which is rep resented as a value of attribute" (membership
func tion). For example, suppose that we have the follow ing tuple t in
a fuzzy relation Employee( Name, Job, Salary, ,,): t = ( John,
Engineer, 60000, 0.67) According to the mathematical definition of
the fuzzy relation, the value 0.67 means only the degree to what
extent this tuple t is possible to belong to the relation.

Database and Expert Systems Applications.pdf

 A requirement-based re-engineering approach should be used when
889.
the information infrastructure of an enterprise is out-of-control and is
not meeting its goals. Typical symptoms are: people can’t share or
integrate data across the enterprise data can’t be combined from
multiple sources the information people need is in a system
somewhere else, not accessible to them the Information System (IS)
staff can’t support the existing infrastructure because of obsolete
platforms or programs designed, built, and supported by one person
the IS department charges are skyrocketing, but people still can’t get
the information they need (what) to conduct. the business; when they
want it, how they want it, or in the form they want it the information
systems don’t support enterprise strategic planning or tactical
decision support information and application requirements have not
been mapped to business functions 643

A Requirement-Based Approach to Data Modeling and Re-

engineering.PDF

In this way, the resulting definition of class is simple and class

890.
operators can be derived easily. Furtl,lermore, since the nested
relational model allows attribute values which can be relations
themselves, our data model supports multivalued properties, which
can be inherited, shared or just local. By adding more superclass and
shared-object attributes, multiple inheritance and multiple object
sharing is supported.

Database and Expert Systems Applications.pdf

891. It is quite obvious that it is no acceptable solution to create a relation

for each class in which each attribute corresponds to one variable
(including the inherited variables) of the class. With such a design the
generalization (an important feature of object oriented modelling)
cannot be correctly expressed in the database, because information
of the general class would have to be stored several times in each of
its subclasses which entails the problems of inconsistency an
redundancy of data. The approach chosen in this paper also requires
a relation for each class declaration.

Database and Expert Systems Applications.pdf

 5 Summary and Outlook In this paper we showed how Web
892.
Intelligence can be used for extracting data automatically from semi-
structured web sites to obtain competitor information for decision
support by using the architecture of the Lixto software. The result of
the process, a structured XML file, can be used by a Business
Intelligence system. We took a closer look at a business scenario for
the company Pirelli, where we described the data upload into Pirelli’s
BI system via an Oracle 9 database.

Lecture Notes in Artificial Intelligence.pdf

893. Sections 3 describes this requirements to meet current and future

business needs. integration taxonomy. Section 4 describes our The
re-engineered design will provide a basis for requirement-based data
modeling approach. Section 5 developing a plan to migrate reusable
applications to the uses an example to illustrate our model integration
B understanding requirements of as-is business processes /data
models and how IS supports users and business . inventory of
baseline requirements implemented in the current IS . composition of
the logical data dimension and tbe physical data dimension of
enterprise data architecture and enterprise data model m integration
of tactical and operational data models among business functions .
integration of strategic, tactical, and operational data models for all
business functions . uniform name for the same data object to
facilitate reuse * same name, meaning, and usage for the sharable
data instances and objects . identification of duplicating or similar
databases for the same business functions l understanding
requirements of as-is business processes /data models and how IS
supports users and business l elimination of obsolete requirements,
modification of changed requirements, and addition of new
requirements l data objects formulated from data usage, processing
rules, attributes, and entities of logical data model l data quality
requirements represented in the re-engineered logical.

A Requirement-Based Approach to Data Modeling and Re-

engineering.PDF

894. Conclusions In this paper, w~ have developed a formal framework for

understanding the connection between a data model supporting
inheritance and object-sharing and the nested relational model. The
translation of inheritance and object-sharing to nested relational
 schemas is shown to be based on inclusion dependencies among
nested relations at different hierarchical levels or different
hierarchies. By using the nested relational model, we have been able
to define a class with superclass and shared-object attributes only
referring to adjacent (in hierarchical terms) classes.

Database and Expert Systems Applications.pdf

895. 1 Introduction In this paper, the requirements for the querying

interface to future environmental information systems are discussed,
and some important concept! for this component are presented. So
far, much of the discussion of EIS has focused on the exten sion of
current database managment systems ( DBMS) in order to cope with
the different types of environmental data. This approach, however,
cannot overcome the conceptual deficien cies of today's DBMS which
were developed originally for the management of bussiness and
administration data, whereBB environmental data has totally different
characteristics, as will be shown below.

Database and Expert Systems Applications.pdf

INTEGRATION OF DATABASE AND HYPERTEXTUAL TECHNOLOGIES

896.
IN DESIGNING A CLINICAL INFORMATION SYSTEM A. CASANOVA
* * Clinica Medica, Universita’ di Cagliari, Via San Giorgio 8,
09124CAGUARI ( ITALY) ABSTRACT This paper presents a Clinical
Information System ( CIS) architecture which has the ability to
integrate different functions performed by autonomous clini cal
structures. The basic idea is the use of hypertex tual technologies, in
order to model an effective management of complex objects (medical
informa tion) which interact in the context of hierarchical structures
(clinical departments). The relational data bases are shown to
provide a reasonable platform to implement the presented
architecture as well as to support medical information retrieval.

Database and Expert Systems Applications.pdf

In the OODB model, each instance (object) has its own identity, and is
897.
associated with a type (or class), which is intuitively a description of
its data structure and its op erational interface. The structural part of
a type (class) can. be recursively constructed by several type
construc tors such as tuple, set, and list. The operational part of a
 type (class) consists of method name, argument types, type of
returned object, and the implementation (code) of the method.

Database and Expert Systems Applications.pdf

898. A Meteorological Database for Numerical and Non-numerical

Processing* Michel Desmarais Alain Leblanc Centre de recherche
informatique de Montreal 3744, rue Jean-Brillant, suite 500 Montreal,
Canada H3T IPI Abstract This paper describes the architecture and
the functionality of the database software developed in Project
Stratus, a project aimed at the development of a knowledge based
system for low clouds forecasting. This read-only software is
composed of a controlling module, DB, and three slave modules.
Each of the slave modules handles a different type of meteorological
data.

Database and Expert Systems Applications.pdf

However, when class which is specified in each instance definition

899.
sentence is not defined, the existing instance object is retrieved from
OMS, or generation of a new instance is conducted. (S4) Selection of
domain D2 The map data that corresponds to each part of the
semantic structure is retrieved from the OMS and is translated into an
expression of a list type, and the analogical inference domain D2 is
determined. The details of the concrete processing contents are as
follows.

Database and Expert Systems Applications.pdf

900. CONCLUSIONS The hypertextual approach revealed itself fit to the

management of a data base on heraldic arms, and allowed the
gathering and the integration of different and connected data. The
implemented system is available on low cost and high diffusion
machines, while the new technology of optical disks has almost no
limits for the development possibilities of the stored information. The
establishment of a thesaurus accompanied by figures, acting with
browsing interface functions, resulted to be an efficient solution, and
therefore represents the means through which it's possible to place
at disposal of the data base manager as well as the user the same
base of knowledge, increasing precision and recall.
 Database and Expert Systems Applications.pdf

Example of select When This expression allows the conditional

901.
execution of a group of actions, based on case logic: when case
condition I then end statemenUistl [case condition2 then
statemenClist2 case conditionN then statemenclistN] [otherwise
statemenUist] In Figure 7, an example of use of when is shown,
integrat ing, at the same time, several of the previous featlll'es we
have presented and some new ones. In particular, in entering several
orders for products, calculating invoices amount and updating class
CLIENT with new orders, invoices and installments, we illustrate the
use of input from screen and the passing of transactions to a method;
the statement sum, which is similar to select; and the creation!
updating of sets by means of the statement union. In the example,
the following data type has been used: ORDER_IN = recordof(
cliencno : oid CLIENT; order: ORDERS; numbecoCinstallments:
integer >= I); method CLIENT -> Princdelinquencpayments( );
actions output cliencnumber, client-> Delinquencpayments( ) to
printer; end; Figure 5.

Database and Expert Systems Applications.pdf

902. The paper presented reports about an ongoing research project

which concentrates on the above mentioned topics. First solutions
are discussed, especially the coupling of ex pert systems and
autonomous databases based on a logical integration of the
heterogeneous data, the combination of probabilistic networks and
production rules to handle uncer tain knowledge and spatial
reasoning, using abstract expres sions with spatial context deriving
from spatial data and op erations. These solutions are implemented
in an expert sys tem prototype, using a typical mine planning task as
background for the research work.

Database and Expert Systems Applications.pdf

A key enabler for the Semantic Web is online onto- logical support for
903.
data, information and knowledge exchange. Given the exponential
growth of the infor- mation available online, automatic processing is
vital to the task of managing and maintaining access to that
information. Used to describe the structure and seman- tics of
information exchange, ontologies are seen to play a key role in areas
 such as knowledge management, B2B e-commerce and other such
burgeoning electronic initiatives.

document ontology2.pdf

904. If this column (with column name of A) was foreign key, and quoted T
j, then establish an object properties identifier and property axiom, in
which, object properties identifier was has_ present column name
(such as has A), property axiom stated that the definition domain of
this object properties was table T i corresponding OWL class, range
was Tj corresponding OWL class; furthermore, if this column was also
primary key, then a class axiom should be. established, which was
used to describe that T i corresponding OWL class was the subclass
of Tj corresponding OWL class; b. If this column was non-foreign key,
then establish a data type properties identifier and property axiom, in
which, object properties identifier was has_present column name,
property axiom indicated that the definition domain of this object
properties was the corresponding OWL class of table T i, range was
the data type that the present column corresponded; furthermore, if
this column was primary key or was with constraint, then the
corresponding constraint should be established.

An Improved Ontology Learning Algorithm from Relational Schema.pdf

method class_name [local var_namel [var_name2 var_ nameN

905.
[preconditions -> method_name ([parameteclist)) [:data_type) ;
data_typel; data_type2; var_namel ===> precondition_list 1
[var_name2 ===> precondition_list2 var_ nameN ===> precondition_
listN]) actions statemenClist end; Figure 2. Method specification
schema 229 An example or select, which uses the class specified in
Figure I, is given in Figure 4, where we use the type: DEL_ INV =
recordof( invoice_number:integer, installmencno:integer);
Other similar sentences for update, delete, insert, min, max, count,
sum, average, and sort in sets are provided.

Database and Expert Systems Applications.pdf

Before we can study agile processes, we must develop the means to

906.
track user activity and analyze it for process-based use. By
developing common process representation protocols for dynamic
process and sharing tools for collecting such representations, we can
 create a shared base for research that will support not only the study
of agile processes in situ, but also enable the development of next-
generation process-oriented knowledge management systems. The
TaskTracer suite offers a set of tools that, when coupled with other
data sources, can gather the data needed for agile process inference
— a first step in understanding process-oriented knowledge
management “in the wild” of users’ environments.

Lecture Notes in Artificial Intelligence.pdf

best practices for system operation (hardware, software etc.) and

907.
support to users 23o Knowledge about the types of software
developed in the organization 25o Organization best practices for
reuse of software items 26o References to technological
developments in Software Engineering 26o Reports on the lessons
learned by the software industry 28o Knowledge about the critical
organizational areas to the achievement of organizational objectives
29o Organization historical data related to product metrics 30o
References. to theoretical knowledge about Software Engineering
31o Knowledge about the organization’s restrictions and about its
weaknesses and strengths 32o Organization historical data related to
quality assurance 33o Organization historical data related to process
metrics 34o Knowledge about the organizational processes 35o
Knowledge about the mission of the organization 36o Organization
best practices for configuration management 37o Organization best
practices for risk management. 38o Knowledge about organizational
guidelines and standards 39o Answers for the most frequent
questions made by organization software developers 40o
Representation of the organizational structure 41o Knowledge about
the organizational processes of client companies 42o Organization
best practices for the process of supplying software products 43o
Representation of the organizational structure of client companies
44o Knowledge about the allocation of client.

Lecture Notes in Artificial Intelligence.pdf

3 Concepts for EIS 3.1 Vague queries and imprecise data In the field
908.
of IR, vague queries and imprecise representations have been
discussed in the context of text retrieval. Wherea~ today's
commercial IR systems are still baeed on simple string search
methods, better representations for text content haw been developed
and tested succesfully in IR research: Stem ming algorithms [35)
 [27) [26) help in searching for different derivations of a word stem,
machine-readable dictionaries [38) and robust parsers support the
identification of noun phrases. Text indexing apptoaches either use a
free vocabulary, that is. every term (single word or phrase) can be
part of a document's description [39) [14]. or they are based on a
controlled vocabu lary, where only index terms from a thesaurus can
be assigned to a document (even if this term does not occur within
the text of the document) [ 28J [15J. In order to cope with the impre
cision of these descriptions, probabilistic IR models have been
developed.

Database and Expert Systems Applications.pdf

The study used: a) prior studies on prevalence of migration; b) profile

909.
and number of returnees from Kingdom of Saudi Arabia; c)ad hoc
data from Regional Bureau of Labour and Social Affairs showing
intensity of irregular migration and related socio-economic problems;
d) youth unemployment as a proxy indicator to select target zones
and woredas from each region. Accordingly, the study selected 12
zones and three sub cities as target areas for the socio economic
assessment. The target zones selected and approved by the TWG for
the socio economic assessment are presented in the table below.

annex_fi_-_inception_phase_report.pdf

Since there is lack of comprehensive data and research about

910.
migration and its determinants, in order to formulate specific
methodologies of intervention for the SINCE Project, the Embassy of
Italy gave ILO the task to undertake comprehensive Socio-Economic
Assessment ( SEA). The SEA has the goal to increase the knowledge
base on migration dynamics, particularly the root causes of
migration, and its nexus with lack of decent livelihood and
employment opportunities. The said assessment constitutes a SINCE
priority action, as its findings provide the basis for developing the
Project’s overall strategies and interventions.

annex_fi_-_inception_phase_report.pdf
.