UNIT‐1 Introduction

Lecture‐1 Motivation: Why data mining?

Lecture‐2 What is data mining?
Lecture‐3 g On what kind of
Data Mining:
data?
Lecture‐4 Data mining functionality
Lecture‐5 Classification of data mining
systems
L
Lecture‐6
6 Major
M j issues
i in
i data
d mining
i i

1
 Unit‐1
Unit 1 Data warehouse and OLAP

L t
Lecture‐7
7 Wh t is
What i a data
d t warehouse?
h ?

Lecture‐8 A multi‐dimensional data model

Lecture‐9 Data warehouse architecture

Lecture‐10&11 Data warehouse implementation

Lecture‐12 From data warehousingg to data miningg

2
 Lecture 1
Lecture‐1
Motivation: Why data mining?

3
 Evolution of Database Technology

• 1960s
1960 and
d earlier:
li
• Data Collection and Database Creation
– Primitive file processing

4
 Evolution of Database Technology

• 1970s ‐ early 1980s:

• Data
D B Base M
Management SSystems
– Hieratical and network database systems
– Relational database Systems
– Query languages: SQL
– Transactions, concurrency control and recovery.
– On
On‐line
line transaction processing (OLTP)

5
 Evolution of Database Technology

• Mid ‐1980s
1980s ‐ present:
– Advanced data models
• Extended relational,
relational object‐relational
object relational
– Advanced application‐oriented DBMS
• spatial,
ti l scientific,
i tifi engineering,
i i temporal,
t l multimedia,
lti di
active, stream and sensor, knowledge‐based

6
 Evolution of Database Technology

• Late 1980s‐present
p
– Advanced Data Analysis
• Data warehouse and OLAP
• Data mining
i i and
d knowledge
k l d discovery
di
• Advanced data mining appliations
• Data mining and socity
• 1990s‐present:
– XML‐based database systems
– Integration with information retrieval
– Data and information integreation

7
 Evolution of Database Technology

• Present – future:
– New generation of integrated data and
information system.
system

8
 Lecture‐2
What Is Data Mining?

9
 What Is Data Mining?

• Data mining refers to extracting or mining

knowledge from large amounts of data.
• Mining of gold from rocks or sand
• Knowledge mining from data, knowledge
extraction,
i data/pattern
d / analysis,
l i data
d
archeology, and data dreding.
• Knowledge Discovery from data, or KDD

10
 Data Mining: A KDD Process

Pattern Evaluation
– Data mining: the core of
knowledge discovery
process
process. Data Mining

Task-relevant Data

Data Warehouse Selection

Data Cleaning

Data Integration

Databases
11
 Steps of a KDD Process
1.
1 Data cleaning
2. Data integration
3
3. Data selection
l i
4. Data transformation
5. Data mining
6
6. Pattern evaluation
7. Knowledge presentaion

12
 Steps
p of a KDD Process
• Learning the application domain:
– relevant
l prior
i kknowledge
l d and d goals
l off
application
• Creating a target data set: data selection
• Data cleaning and preprocessing
• Data reduction and transformation:
– Find useful features, dimensionality/variable
reduction, invariant representation.

13
 Steps of a KDD Process
• Choosing functions of data mining
– summarization, classification, regression, association,
clustering.
• Choosing the mining algorithms
• Data mining: search for patterns of interest
• Pattern evaluation and knowledge presentation
– visualization, transformation, removing redundant
patterns, etc.
• Use of discovered knowledge

14
Architecture of a Typical Data
Mi i System
Mining S t
G hi l user interface
Graphical i f

Pattern evaluation

Data
a a mining
g engine
g
Knowledge-base
Database or data
warehouse server
Data cleaning & data integration Filtering

Data
Databases Warehouse

15
Data Mining and Business Intelligence
Increasing potential
to support
business decisions End User
Making
Decisions

Data Presentation Business

Analyst
Visualization Techniques
Data
D t Mi
Mining
i Data
D t
Information Discovery Analyst

Data Exploration
Statistical Analysis, Querying and Reporting

Data Warehouses / Data Marts

OLAP, MDA DBA
Data Sources
Paper, Files, Information Providers, Database Systems, OLTP

16
 Lecture‐3
Data Mining: On What Kind of Data?

17
 Data Mining: On What Kind of Data?

• Relational databases
• Data warehouses
• Transactional databases

18
 Data Mining: On What Kind of Data?

• Advanced DB and information repositories

– Object‐oriented and object‐relational databases
– Spatial databases
– Time‐series data and temporal data
– Text
T t databases
d t b andd multimedia
lti di databases
d t b
– Heterogeneous and legacy databases
– WWW

19
 Lecture‐4
Lecture 4
Data Mining Functionalities

20
 Data Miningg Functionalities

• Concept
p description:
p Characterization and
discrimination
– Data can be associated with classes or concepts
p
– Ex. AllElectronics store classes of items for sale include
computer and printers.
– Description of class or concept called class/concept
description.
– Data characterization
– Data discrimination

21
 Data Mining Functionalities
• Mining Frequent Patterns
Patterns, Associations
Associations, and
Correlations
• Frequent patters
patters‐ patterns occurs frequently
• Item sets, subsequences and substructures
• Frequent item set
• Sequential patterns
• Structured patterns

22
 Data Mining Functionalities
• Association Analysis
– Multi‐dimensional vs. single‐dimensional
association
– age(X, “20..29”) ^ income(X, “20..29K”) => buys(X,
“PC”) [[supportt = 2%
2%, confidence
fid = 60%]
– contains(T, “computer”) => contains(x,
“ f
“software”)
”) [[support=1%, confidence=75%]
f ]

23
 Data Mining Functionalities

• Classification and Prediction

– Finding models (functions) that describe and
distinguish data classes or concepts for predict the
class whose label is unknown
– E.g., classify countries based on climate, or classify
cars based on gas mileage
– Models: decision‐tree, classification rules (if‐then),
neurall network
t k
– Prediction: Predict some unknown or missing
numerical values
24
 Data Mining Functionalities

• Cluster analysis
– Analyze class‐labeled data objects, clustering
analyze data objects without consulting a known
class label.
– Clustering
Cl t i b basedd on th
the principle:
i i l maximizing
i i i ththe
intra‐class similarity and minimizing the interclass
similarity

25
 Data Miningg Functionalities
• Outlier analysis
– Outlier: a data object that does not comply with the general behavior
of the model of the data
– It can be considered as noise or exception but is quite useful in fraud
detection, rare events analysis

• Trend and evolution analysis

y
– Trend and deviation: regression analysis
– Sequential pattern mining,
mining periodicity analysis
– Similarity‐based analysis

26
 Lecture‐5
Lecture 5
Data Mining: Classification Schemes

27
 Data Mining: Confluence of Multiple
Disciplines
Database
Statistics
Technology

Information
Science Data Mining MachineLearning

Visualization Other
Disciplines

28
 Data Mining: Classification Schemes

• General functionalityy
– Descriptive data mining
– Predictive data mining

• Data mining various criteria's:

– Kinds of databases to be mined
– Kinds of knowledge to be discovered
– Kinds of techniques utilized
– Kinds of applications
pp adapted
p

29
 Data Mining: Classification Schemes
• Databases to be mined
– Relational,, transactional,, object‐oriented,
j , object‐
j
relational, active, spatial, time‐series, text, multi‐media,
heterogeneous, legacy, WWW, etc.
• Knowledge to be mined
– Characterization, discrimination, association,
classification, clustering, trend, deviation and outlier
analysis etc.
analysis, etc
– Multiple/integrated functions and mining at multiple
levels
• analysis, Web mining, Weblog analysis, etc.

30
 Data Mining: Classification Schemes

• Techniques utilized
– Database‐oriented, data warehouse (OLAP),
machine learning
learning, statistics
statistics, visualization,
visualization
neural network, etc.
• Applications
A li i adapted
d d
– Retail, telecommunication, banking, fraud
analysis, DNA mining, stock market

31
 Lecture‐6
Lecture 6
Major Issues in Data Mining

32
 Major Issues in Data Mining

• Mining methodology and user interaction issues

– Mining different kinds of knowledge in databases
– Interactive mining of knowledge at multiple levels of
abstraction
– Incorporation of background knowledge
– Data mining query languages and ad‐hoc data mining
– Expression and visualization of data mining results
– Handling noise and incomplete data
– Pattern evaluation: the interestingness problem

33
 Major Issues in Data Mining
• Performance issues

– Efficiency
Effi i and
d scalability
l bilit off data
d t mining
i i algorithms
l ith
– Parallel, distributed and incremental mining
methods
h d

34
 Major Issues in Data Mining

• Issues relatingg to the diversityy of data types

yp

– Handling relational and complex types of data

– Mining
Minin information from hetero
heterogeneous
eneo s databases
and global information systems (WWW)

35
 Lecture‐7

Wh t iis Data
What D t WWarehouse?
h ?

36
 What is Data Warehouse?
• Defined in many different ways
– A decision support database that is maintained separately
from the organization’s operational database
– Support information processing by providing a solid
platform of consolidated, historical data for analysis.

• “A data warehouse is a subject‐oriented, integrated, time‐

variant and nonvolatile collection of data in support of
variant,
management’s decision‐making process.”—W. H. Inmon

• Data warehousing:
h i
– The process of constructing and using data warehouses

37
 D t Warehouse—Subject‐Oriented
Data W h S bj t O i t d
• Organized around major subjects
subjects, such as customer,
customer product,
product
sales.
• Focusing on the modeling and analysis of data for decision
makers, not on daily operations or transaction processing.
• Provide a simple and concise view around particular subject
issues by excluding data that are not useful in the decision
support process.
process

38
 Data Warehouse—Integrated
• Constructed by integrating multiple, heterogeneous
data sources
– relational databases, flat files, on‐line transaction records
• Data cleaningg and data integration
g techniques
q are
applied.
– Ensure consistency in naming conventions, encoding
structures, attribute measures, etc. among different data
sources
• E.g.,
g , Hotel price:
p currency,
y, tax,, breakfast covered,, etc.
– When data is moved to the warehouse, it is converted.

39
 Data Warehouse—Time
Warehouse Time Variant
• The time horizon for the data warehouse is
significantly longer than that of operational systems.
– Operational database: current value data.
– Data warehouse data: provide information from a historical
perspective (e.g., past 5‐10 years)
• Every key
k structure iin the
h ddata warehouse
h
– Contains an element of time, explicitly or implicitly
– But the key of operational data may or may not contain
“time element”.

40
 Data Warehouse
Warehouse—Non‐Volatile
Non Volatile
• A physically separate store of data transformed from
the operational environment.
• Operational update of data does not occur in the
data warehouse environment.
– Does not require transaction processing, recovery, and
concurrencyy control mechanisms
– Requires only two operations in data accessing:
• initial
t a loading
oad g of data aand
d access of data
data.

41
 Data Warehouse vs
vs. Operational DBMS
• Distinct features (OLTP vs. OLAP):
– User
U andd system
t orientation:
i t ti customer
t vs. market
k t
– Data contents: current, detailed vs. historical, consolidated
– Database design: ER + application vs.
vs star + subject
– View: current, local vs. evolutionary, integrated
– Access patterns: update vs
vs. read
read‐only
only but complex queries

42
 Data Warehouse vs. Operational DBMS
• OLTP (on‐line transaction processing)
– Major task of traditional relational DBMS
– Day‐to‐day operations: purchasing, inventory, banking,
manufacturing, payroll, registration, accounting, etc.

• OLAP (on‐line analytical processing)

– Major task of data warehouse system
– Data analysis and decision making

43
 OLTP vs
vs. OLAP
OLTP OLAP
users clerk, IT professional knowledge worker
function day to day operations decision support
DB design
g application-oriented
pp subject-oriented
j
data current, up-to-date historical,
detailed, flat relational summarized, multidimensional
isolated integrated, consolidated
usage repetitive ad hoc
ad-hoc
access read/write lots of scans
index/hash on prim. key
unit of work short, simple transaction complex queryy
# records accessed tens millions
#users thousands hundreds
DB size 100MB-GB 100GB-TB
metric transaction throughput query throughput, response

44
 Why Separate Data Warehouse?
• High performance for both systems

– DBMS— tuned for OLTP: access methods, indexing,

concurrency control, recovery

– Warehouse—tuned
W h t d for
f OLAP:
OLAP complex
l OLAP
queries, multidimensional view, consolidation.

45
 Why Separate Data Warehouse?
• Different functions and different data:
– missing data: Decision support requires historical
data which operational DBs do not typically
maintain
– data consolidation: DS requires consolidation
(aggregation summari
(aggregation, summarization)
ation) of data from
heterogeneous sources
– data quality: different sources typically use
inconsistent data representations, codes and
formats which have to be reconciled

46
 L
Lecture‐8
8

A multi‐dimensional data model

47
 Cube: A Lattice of Cuboids

all
0-D(apex) cuboid

time item location supplier

1-D cuboids

time item
time,item time location
time,location item location
item,location location supplier
location,supplier
2-D cuboids
time,supplier item,supplier

time,location,supplier
time,item,location 3-D cuboids
time,item,supplier item,location,supplier

4-D(base) cuboid
time, item, location, supplier

48
Conceptual Modeling of Data Warehouses

• Modeling data warehouses: dimensions & measures

– Star schema: A fact table in the middle connected to a set of
di
dimension
i tables
bl
– Snowflake schema: A refinement of star schema where
some dimensional hierarchy is normalized into a set of
smaller dimension tables, forming a shape similar to
snowflake
– Fact constellations: Multiple fact tables share dimension
tables, viewed as a collection of stars, therefore called
galaxy schema or fact constellation

49
 Example of Star Schema
time
time_key
time key item
day item_key
day_of_the_week Sales Fact Table item_name
month brand
quarter time key
time_key type
year supplier_type
item_key
branch key
branch_key
branch location
location_key
branch_key location_key
bbranch
a c _name
a e units sold
units_sold street
branch_type city
dollars_sold province_or_street
country
avg_sales
Measures

50
 Example of Snowflake Schema
time
time_key
time key item
day item_key supplier
day_of_the_week Sales Fact Table item_name supplier_key
month brand supplier_type
quarter time key
time_key t
type
year item_key supplier_key

branch_key
y
branch location
location_key
location_key
branch_key
units_sold street
bbranch
a c _name
a e
city_key
it k
branch_type city
dollars_sold
city_key
avg_sales cityy
Measures province_or_street
country

51
 Example
p of Fact Constellation
time
time_key item Shipping Fact Table
day item_key
day_of_the_week Sales Fact Table item_name time_key
month brand
quarter time_keyy type
yp item_key
year supplier_type shipper_key
item_key
branch_key from_location

branch location_key location to_location

branch_key location_key dollars_cost
branch_name
units_sold
street
branch_type dollars_sold city units_shipped
province_or_street
avg_sales country shipper
M
Measures shipper_key
shipper_name
location_key
shipper_type
52
A Data Mining Query Language, DMQL: Language
Primitives

• Cube Definition (Fact Table)

define cube <cube_name> [<dimension_list>]:
<measure_list>
• Dimension Definition ( Dimension Table )
define dimension <dimension_name> as
(
(<attribute_or_subdimension_list>)
ib bdi i li )
• Special Case (Shared Dimension Tables)
– Fi
Firstt ti
time as “cube
“ b definition”
d fi iti ”
– define dimension <dimension_name> as
<dimension_name_first_time> in cube
<cube_name_first_time>

53
 Defining a Star Schema in DMQL

define cube sales_star [[time,, item,, branch,, location]:

]
dollars_sold = sum(sales_in_dollars), avg_sales =
avg(sales_in_dollars), units_sold = count(*)
define dimension time as (time_key, day, day_of_week, month,
quarter, year)
define dimension item as (item
(item_key,
key item
item_name,
name brand
brand, type
type,
supplier_type)
define dimension branch as (branch_key, branch_name,
branch_type)
define dimension location as (location_key, street, city,
province or state country)
province_or_state,

54
 Defining a Snowflake Schema in DMQL

define cube sales_snowflake [time, item, branch, location]:

dollars_sold
d ll ld = sum(sales_in_dollars),
( l i d ll ) avg_sales l =
avg(sales_in_dollars), units_sold = count(*)
define dimension time as (time
(time_key,
key day
day, day
day_of_week,
of week
month, quarter, year)
define dimension item as (item_key,
(item key item_name,
item name brand,
brand
type, supplier(supplier_key, supplier_type))

55
 Defining a Snowflake Schema in DMQL

define dimension branch as (branch

(branch_key,
key
branch_name, branch_type)
define dimension location as (location_key,
(
street, city(city_key, province_or_state,
country))

56
 Defining a Fact Constellation in DMQL
define cube sales [time, item, branch, location]:
dollars_sold = sum(sales_in_dollars), avg_sales =
avg(sales_in_dollars),
( ) units_sold = count(*)
(*)
define dimension time as (time_key, day, day_of_week, month,
quarter,, year)
q y )
define dimension item as (item_key, item_name, brand, type,
supplier_type)
define dimension branch as (branch_key,
(branch key branch_name,
branch name branch_type)
branch type)
define dimension location as (location_key, street, city,
province_or_state, country)

57
 Defining a Fact Constellation in DMQL
define cube shipping [time, item, shipper, from_location,
to_location]:
dollar_cost = sum(cost_in_dollars), unit_shipped =
count( )
count(*)
define dimension time as time in cube sales
define dimension item as item in cube sales
define dimension shipper as (shipper_key, shipper_name,
location as location in cube sales, shipper_type)
definee dimension
de d e s o from o _location
ocat o as location
ocat o in cube sa
sales
es
define dimension to_location as location in cube sales

58
 Measures: Three Categories
• distributive: if the result derived by applying the
function to n aggregate values is the same as that
derived by applying the function on all the data
without partitioning
partitioning.
• E.g., count(), sum(), min(), max().
• algebraic: if it can be computed by an algebraic
function with M arguments (where M is a bounded
integer) each of which is obtained by applying a
integer),
distributive aggregate function.
• E.g.,
E g avg(),
avg() min
min_N(),
N() standard_deviation().
standard deviation()

59
 Measures: Three Categories

• holistic: if there is no constant bound on the

storage size needed to describe a sub
aggregate.
aggregate
• E.g., median(), mode(), rank().

60
 A Concept Hierarchy: Dimension (location)

all all

region Europe ... North_America

country Germany ... Spain Canada ... Mexico

city Frankfurt ... Vancouver ... Toronto

office L. Chan ... M. Wind

61
 M ltidi
Multidimensional
i l Data
D t
• Sales volume as a function of product,
month, and region Dimensions: Product, Location, Time
Hierarchical summarization paths

Industry Region Year

Category Country Quarter

Prooduct

Product City Month Week

Office Day

Month

62
 A Sample Data Cube
Date Total annual sales
1Qtr 2Qtr 3Qtr 4Qt
4Qtr sum of TV in U
U.S.A.
SA
TV
PC U.S.A
VCR

Country
y
sum
Canada

C
Mexico

sum

63
Cuboids Corresponding to the Cube

all
0-D(apex) cuboid
product
d date country
1-D cuboids

pproduct,date
, pproduct,country
, y date,, country
y
2-D cuboids

3 D(b ) cuboid
3-D(base) b id
product, date, country

64
 OLAP Operations
p

• Roll up (drill‐up):
(drill up): summarize data
– by climbing up hierarchy or by
dimension reduction
• Drill down (roll down): reverse of roll‐up
– from
f higher
hi h level
l l summary to
t lower
l
level summary or detailed data, or
introducing new dimensions
• Slice and dice:
– project and select

65
 OLAP Operations
• Pivot (rotate):
– reorient the cube, visualization, 3D to
series
se es of
o 2D planes.
p a es
• Other operations
– drill across: involving (across) more
than one fact table
– drill through: through the bottom level
of the cube to its back-end relational
tables (using SQL)

66
 Lecture‐9

Data warehouse architecture

67
 Steps for the Design and Construction of
Data Warehouse
h

• The design of a data warehouse: a business

analysis framework
• The process of data warehouse design
• A three‐tier data ware house architecture

68
Design of a Data Warehouse: A Business Analysis
Framework

• Four
F views
i regarding
di ththe d
design
i off a d
data
t warehouse
h
– Top‐down view
• allows selection of the relevant information
necessary for the data warehouse

69
Design of a Data Warehouse: A Business Analysis
F
Frameworkk
– Data warehouse view
• consists of fact tables and dimension tables

– Data source view

• exposes the information being captured,
captured stored,
stored and
managed by operational systems

– Business query view

• sees the perspectives

70
 Data Warehouse Design Process

• Top‐down,
Top down, bottom
bottom‐up
up approaches or a combination
of both
– Top‐down: Starts with overall design and planning
(mature)
– Bottom‐up: Starts with experiments and prototypes (rapid)
• From software engineering point of view
– Waterfall: structured and systematic analysis at each step
before proceeding to the next
– Spiral: rapid generation of increasingly functional systems,
short turn around time, quick turn around

71
 Data Warehouse Design Process

• Typical data warehouse design process

– Choose a business process to model, e.g., orders,
invoices etc.
invoices, etc
– Choose the grain (atomic level of data) of the
business process
– Choose the dimensions that will apply to each fact
table record
– Choose the measure that will populate each fact
table record

72
 Multi--Tiered Architecture
Multi

Monitor
& OLAP Server
other Metadata
Integrator
sources
Analysis
A l i
Operational Extract Query
Transform Data Serve Reports
DBs
Load
Refresh
Warehouse Data mining

Data Marts

Data Sources Data Storage OLAP Engine Front-End Tools

73
 Metadata Repository
p y
• Meta data is the data defining warehouse objects. It has the
f ll i kinds
following ki d
– Description of the structure of the warehouse
• schema, view, dimensions, hierarchies, derived data defn, data mart locations
and contents
– Operational meta‐data
• data lineage (history of migrated data and transformation path), currency of
data (active, archived, or purged), monitoring information (warehouse usage
statistics, error reports, audit trails)
– The algorithms used for summarization
– The mapping from operational environment to the data warehouse
– Data related to system performance
• warehouse schema, view and derived data definitions
– Business data
• business terms and definitions, ownership of data, charging policies

74
 Data Warehouse Back‐End Tools and Utilities

• Data extraction:
– get data from multiple, heterogeneous, and external
sources
• Data cleaning:
– detect errors in the data and rectify them when possible
• Data transformation:
– convert data from legacy or host format to warehouse
format
• Load:
– sort, summarize, consolidate, compute views, check
integrity, and build indices and partitions
• Refresh
– propagate the updates from the data sources to the
warehouse

75
 Three Data Warehouse Models
• Enterprise warehouse
– collects all of the information about subjects
j spanning
p g the entire
organization
• Data Mart
– a subset
b off corporate‐wide id data
d that
h isi off value
l to a specific
ifi groups
of users. Its scope is confined to specific, selected groups, such as
marketing data mart
• Independent vs. dependent (directly from warehouse) data
mart
• Virtual
Vi t l warehouse
h
– A set of views over operational databases
– Onlyy some of the p possible summaryy views mayy be materialized

76
 Data Warehouse Development: A
Recommended Approach

Multi-Tier Data
Warehouse
Distributed
Data Marts

Enterprise
E t i
Data Data
Data
Mart Mart
Warehouse

Model refinement Model refinement

Define a high-level corporate data model

77
 Types of OLAP Servers
• Relational OLAP (ROLAP)
– Use relational or extended‐relational DBMS to store and
manage warehouse data and OLAP middle ware to support
missing pieces
– Include optimization of DBMS backend,
backend implementation of
aggregation navigation logic, and additional tools and
services
– greater scalability
• Multidimensional OLAP (MOLAP)
– Array‐based multidimensional storage engine (sparse matrix
techniques)
– fast indexing to pre‐computed
pre computed summarized data

78
 Types of OLAP Servers
• Hybrid OLAP (HOLAP)
– User flexibility, e.g., low level: relational, high‐
level: array
• Specialized SQL servers
– specialized support for SQL queries over
star/snowflake schemas

79
 Lecture‐10
Lecture 10 & 11

Data warehouse implementation

80
 Efficient Data Cube Computation
• Data cube can be viewed as a lattice of cuboids
– The
Th bbottom‐most
tt t cuboid
b id is
i the
th base
b cuboid
b id
– The top‐most cuboid (apex) contains only one cell
– How many cuboids in an n‐dimensional
n dimensional cube with L levels?

n
• T  of( Ldata
Materialization i  1) cube
i 1
– Materialize every (cuboid) (full materialization), none (no
materialization) or some (partial materialization)
materialization),
– Selection of which cuboids to materialize
• Based on size,, sharing,
g, access frequency,
q y, etc.

81
 Cube Operation
• Cube definition and computation in DMQL
define cube sales[item, city, year]: sum(sales_in_dollars)
compute cube sales
• Transform it into a SQL‐like language (with a new operator cube
by introduced by Gray et al.
by, al ’96)
96)
SELECT item, city, year, SUM (amount)
()
FROM SALES
CUBE BY item, city, year
(city) (item) (year)
• Need compute the following Group‐Bys
(date, product,
(date product customer),
customer)
(date,product),(date, customer), (product, customer),
(date), (product), (customer) (city, item) (city, year) (item, year)
()
(city, item, year)

82
Cube Computation: ROLAP‐Based Method

• Efficient cube computation

p methods
– ROLAP‐based cubing algorithms (Agarwal et al’96)
– Array‐based cubing algorithm (Zhao et al’97)
– Bottom‐up
Bottom up computation method (Bayer & Ramarkrishnan’99)
Ramarkrishnan 99)

• ROLAP‐based cubing algorithms

– SSorting,
ti h hashing,
hi andd grouping
i operations
ti are applied
li d to
t th
the
dimension attributes in order to reorder and cluster related tuples
– Grouping is performed on some sub aggregates as a “partial
grouping step”
– Aggregates may be computed from previously computed
aggregates,
t rather
th than
th from
f the
th base
b fact
f t table
t bl

83
 Multi‐way Array Aggregation for
Cube Computation
• Partition arrays into chunks (a small sub cube which fits in
memory).
• Compressed sparse array addressing: (chunk_id, offset)
• Compute aggregates in “multi way” by visiting cube cells in the
order which minimizes the # of times to visit each cell, and
reduces memory access and storage cost.
cost

84
Multi‐way Array Aggregation for
C b C
Cube Computation
t ti

C c3 61
c2 45
62 63 64
46 47 48
c11 29 30 31 32
c0
B13 14 15 16 60
b3 44
B b2 28 56
9
40
24 52
b1 5
36
20
b0 1 2 3 4
a0 a1 a2 a3
A

85
 Multi‐Way Array Aggregation for Cube
Computation
• Method: the planes should be sorted and
computed according to their size in ascending
order.
– Idea: keep the smallest plane in the main
memory, fetch and compute only one chunk at a
time for the largest plane
• Limitation of the method: computing well
only for a small number of dimensions
– If there are a large number of dimensions,
“bottom‐up computation” and iceberg cube
computation methods can be explored

86
 Indexing OLAP Data: Bitmap Index
• Index on a particular column
• Each value in the column has a bit vector: bit‐op is fast
• The length of the bit vector: # of records in the base table
• The i‐th bit is set if the i‐th row of the base table has the
value for the indexed column
• not suitable for high cardinality domains

Base table Index on Region Index on Type

Cust Region Type RecIDAsia Europe America RecID Retail Dealer
C1 A i
Asia R t il
Retail 1 1 0 0 1 1 0
C2 Europe Dealer 2 0 1 0 2 0 1
C3 Asia Dealer 3 1 0 0 3 0 1
C4 A
America
i R t il
Retail 4 0 0 1 4 1 0
C5 Europe Dealer 5 0 1 0 5 0 1

87
 Indexing OLAP Data: Join Indices
• Join index: JI(R‐id, S‐id) where R (R‐id, …) 
S (S‐id, …)
• Traditional indices map p the values to a list of
record ids
– It materializes relational join in JI file and speeds
upp relational jjoin — a rather costlyy operation
p
• In data warehouses, join index relates the
values of the dimensions of a start schema
to rows in the fact table.
– E.g. fact table: Sales and two dimensions city and
product
• A join index on city maintains for each distinct
city a list of R‐IDs of the tuples recording the
Sales in the city
– Join indices can span multiple dimensions

88
 Efficient Processing OLAP Queries
• Determine which operations should be performed on
the available cuboids:
– transform drill, roll, etc. into corresponding SQL and/or OLAP
operations,
i e.g, dice
di = selection
l i + projection
j i

• Determine to which materialized cuboid(s) the relevant

operations should be applied.
• Exploring
E l i indexing
i d i structures
t t and
d compressed
d vs.
dense array structures in MOLAP

89
 Lecture‐12

From data warehousing to data

mining
i i

90
 Data Warehouse Usage
• Three kinds of data warehouse applications
– Information processing
• supports querying, basic statistical analysis, and reporting using
crosstabs, tables, charts and graphs
– Analytical
l l processing
• multidimensional analysis of data warehouse data
pp
• supports basic OLAP operations,
p , slice‐dice,, drilling,
g, pivoting
p g
– Data mining
• knowledge discovery from hidden patterns
• supports associations, constructing analytical models,
performing classification and prediction, and presenting the
mining results using visualization tools.
• Differences among the three tasks

91
From On‐Line Analytical Processing to On Line Analytical
Mining (OLAM)

• Why online analytical mining?

– High quality of data in data warehouses
• DW contains integrated, consistent, cleaned data
– Available
l bl information
f processing structure surrounding
d data
d
warehouses
• ODBC, OLEDB, Web accessing, service facilities, reporting and
O
OLAP tools
l
– OLAP‐based exploratory data analysis
• mining with drilling, dicing, pivoting, etc.
– On‐line selection of data mining functions
• integration and swapping of multiple mining functions,
algorithms,
g , and tasks.
• Architecture of OLAM
92
 An OLAM Architecture
Mi i query
Mining Mi i result
Mining l L
Layer4
4
User Interface
User GUI API
Layer3
OLAM OLAP
Engine
g Engine
g OLAP/OLAM

Data Cube API

Layer2
MDDB
MDDB
Meta Data

Filtering&Integration Database API Filtering

y
Layer1
Data cleaning Data
Databases Data
Data integration
Warehouse
Repository
93