Data Mining:

Concepts and Techniques

—Lecture 1 —

MOI UNIVERSITY
DEPARTMENT OF IT
SCHOOL OF INFORMATION SCIENCES

1
 Introduction
 Why Data Mining?
 What Is Data Mining?
 A Multi-Dimensional View of Data Mining
 What Kind of Data Can Be Mined?
 What Kinds of Patterns Can Be Mined?
 What Technology Are Used?
 What Kind of Applications Are Targeted?
 Major Issues in Data Mining
 A Brief History of Data Mining and Data Mining Society
 Summary

2
 Why Data Mining?
 The Explosive Growth of Data: from terabytes to petabytes
 Data collection and data availability

Automated data collection tools, database systems, Web,
computerized society
 Major sources of abundant data

Business: Web, e-commerce, transactions, stocks, …

Science: Remote sensing, bioinformatics, scientific
simulation, …

Society and everyone: news, digital cameras, YouTube
 We are drowning in data, but starving for knowledge!
 “Necessity is the mother of invention”—Data mining—
Automated analysis of massive data sets
3
 Evolution of Sciences
 Before 1600, empirical science
 1600-1950s, theoretical science
 Each discipline has grown a theoretical component. Theoretical models often
motivate experiments and generalize our understanding.
 1950s-1990s, computational science
 Over the last 50 years, most disciplines have grown a third, computational
branch (e.g. empirical, theoretical, and computational ecology, or physics, or
linguistics.)
 Computational Science traditionally meant simulation. It grew out of our
inability to find closed-form solutions for complex mathematical models.
 1990-now, data science
 The flood of data from new scientific instruments and simulations
 The ability to economically store and manage petabytes of data online
 The Internet and computing Grid that makes all these archives universally
accessible
 Scientific info. management, acquisition, organization, query, and visualization
tasks scale almost linearly with data volumes. Data mining is a major new
challenge!
 Jim Gray and Alex Szalay, The World Wide Telescope: An Archetype for Online
Science, Comm. ACM, 45(11): 50-54, Nov. 2002
4
 Evolution of Database Technology
 1960s:
 Data collection, database creation, IMS and network DBMS
 1970s:
 Relational data model, relational DBMS implementation
 1980s:
 RDBMS, advanced data models (extended-relational, OO, deductive,
etc.)
 Application-oriented DBMS (spatial, scientific, engineering, etc.)
 1990s:
 Data mining, data warehousing, multimedia databases, and Web
databases
 2000s
 Stream data management and mining
 Data mining and its applications
 Web technology (XML, data integration) and global information systems

5
 What Is Data Mining?

 Data mining (knowledge discovery from data)

 Extraction of interesting (non-trivial, implicit, previously
unknown and potentially useful) patterns or knowledge
from huge amount of data
 Data mining: a misnomer?
 Alternative names
 Knowledge discovery (mining) in databases (KDD),
knowledge extraction, data/pattern analysis, data
archeology, data dredging, information harvesting,
business intelligence, etc.
 Watch out: Is everything “data mining”?
 Simple search and query processing
 (Deductive) expert systems
7
 Knowledge Discovery (KDD) Process
Knowledge
 This is a view from typical
database systems and data
Pattern Evaluation
warehousing communities
 Data mining plays an essential
role in the knowledge discovery
process Data Mining

Task-relevant Data

Data Selection
Warehouse
Data Cleaning

Data Integration

Databases
8
 Example: A Web Mining Framework

 Web mining usually involves

 Data cleaning
 Data integration from multiple sources
 Warehousing the data
 Data cube construction
 Data selection for data mining
 Data mining
 Presentation of the mining results
 Patterns and knowledge to be used or stored into
knowledge-base

9
 Data Mining in Business Intelligence

Increasing potential
to support
business decisions End User
Decision
Making

Data Presentation Business

Analyst
Visualization Techniques
Data Mining Data
Information Discovery Analyst

Data Exploration
Statistical Summary, Querying, and Reporting

Data Preprocessing/Integration, Data Warehouses

DBA
Data Sources
Paper, Files, Web documents, Scientific experiments, Database Systems
10
Example: Mining vs. Data Exploration
 Business intelligence view
 Warehouse, data cube, reporting but not much
mining
 Business objects vs. data mining tools
 Supply chain example: tools
 Data presentation
 Exploration

11
KDD Process: A Typical View from ML and Statistics

Patt
Info ern
rma
Input Data Data Pre- Data Post- t io
Processing Mining Processing n
Kno
wled
ge

Data integration Pattern discovery Pattern evaluation

Normalization Association & Pattern selection
correlation
Feature selection Classification Pattern interpretation
Dimension reduction Clustering Pattern visualization
Outlier analysis
…………

 This is a view from typical machine learning and statistics communities

12
 Example: Medical Data Mining
 Health care & medical data mining – often
adopted such a view in statistics and
machine learning
 Preprocessing of the data (including feature
extraction and dimension reduction)
 Classification or/and clustering processes
 Post-processing for presentation

13
 Multi-Dimensional View of Data Mining
 Data to be mined
 Database data (extended-relational, object-oriented,

heterogeneous, legacy), data warehouse, transactional data,

stream, spatiotemporal, time-series, sequence, text and web,
multi-media, graphs & social and information networks
 Knowledge to be mined (or: Data mining functions)
 Characterization, discrimination, association, classification,

clustering, trend/deviation, outlier analysis, etc.

 Descriptive vs. predictive data mining

 Multiple/integrated functions and mining at multiple levels

 Techniques utilized
 Data-intensive, data warehouse (OLAP), machine learning,

statistics, pattern recognition, visualization, high-performance,

etc.
 Applications adapted
 Retail, telecommunication, banking, fraud analysis, bio-data

mining, stock market analysis, text mining, Web mining, etc. 15

 Data Mining: On What Kinds of Data?
 Database-oriented data sets and applications
 Relational database, data warehouse, transactional database
 Advanced data sets and advanced applications
 Data streams and sensor data
 Time-series data, temporal data, sequence data (incl. bio-
sequences)
 Structure data, graphs, social networks and multi-linked data
 Object-relational databases
 Heterogeneous databases and legacy databases
 Spatial data and spatiotemporal data
 Multimedia database
 Text databases
 The World-Wide Web
17
 Data Mining Function: (1) Generalization

 Information integration and data warehouse

construction
 Data cleaning, transformation, integration, and
multidimensional data model
 Data cube technology
 Scalable methods for computing (i.e.,
materializing) multidimensional aggregates
 OLAP (online analytical processing)
 Multidimensional concept description:
Characterization and discrimination
 Generalize, summarize, and contrast data
characteristics, e.g., dry vs. wet region
19
 Data Mining Function: (2) Association and
Correlation Analysis
 Frequent patterns (or frequent itemsets)
 What items are frequently purchased together
in your Walmart?
 Association, correlation vs. causality
 A typical association rule

Diaper  Beer [0.5%, 75%] (support, confidence)
 Are strongly associated items also strongly
correlated?
 How to mine such patterns and rules efficiently in
large datasets?
 How to use such patterns for classification,
clustering, and other applications?
20
 Data Mining Function: (3) Classification

 Classification and label prediction


Construct models (functions) based on some training
examples

Describe and distinguish classes or concepts for future
prediction

E.g., classify countries based on (climate), or classify cars
based on (gas mileage)

Predict some unknown class labels
 Typical methods

Decision trees, naïve Bayesian classification, support vector
machines, neural networks, rule-based classification, pattern-
based classification, logistic regression, …
 Typical applications:

Credit card fraud detection, direct marketing, classifying
stars, diseases, web-pages, …
21
Data Mining Function: (4) Cluster Analysis

 Unsupervised learning (i.e., Class label is unknown)

 Group data to form new categories (i.e., clusters),
e.g., cluster houses to find distribution patterns
 Principle: Maximizing intra-class similarity &
minimizing interclass similarity
 Many methods and applications

22
Data Mining Function: (5) Outlier Analysis
 Outlier analysis
 Outlier: A data object that does not comply with the general
behavior of the data
 Noise or exception? ― One person’s garbage could be
another person’s treasure
 Methods: by product of clustering or regression analysis, …
 Useful in fraud detection, rare events analysis

23
Time and Ordering: Sequential Pattern, Trend and
Evolution Analysis

 Sequence, trend and evolution analysis

 Trend, time-series, and deviation analysis: e.g.,

regression and value prediction

 Sequential pattern mining


e.g., first buy digital camera, then buy large SD memory cards
 Periodicity analysis
 Motifs and biological sequence analysis

Approximate and consecutive motifs
 Similarity-based analysis
 Mining data streams
 Ordered, time-varying, potentially infinite, data

streams

24
 Structure and Network Analysis
 Graph mining
 Finding frequent subgraphs (e.g., chemical compounds), trees

(XML), substructures (web fragments)

 Information network analysis
 Social networks: actors (objects, nodes) and relationships

(edges)

e.g., author networks in CS, terrorist networks
 Multiple heterogeneous networks


A person could be multiple information networks: friends,
family, classmates, …
 Links carry a lot of semantic information: Link mining

 Web mining
 Web is a big information network: from PageRank to Google

 Analysis of Web information networks


Web community discovery, opinion mining, usage mining,
…
25
 Evaluation of Knowledge
 Are all mined knowledge interesting?
 One can mine tremendous amount of “patterns” and
knowledge
 Some may fit only certain dimension space (time, location, …)
 Some may not be representative, may be transient, …
 Evaluation of mined knowledge → directly mine only
interesting knowledge?
 Descriptive vs. predictive
 Coverage
 Typicality vs. novelty
 Accuracy
 Timeliness
 …
26
Data Mining: Confluence of Multiple Disciplines

Machine Pattern Statistics

Learning Recognition

Applications Data Mining Visualization

Algorithm Database High-Performance

Technology Computing

28
 Why Confluence of Multiple Disciplines?
 Tremendous amount of data
 Algorithms must be highly scalable to handle such as tera-
bytes of data
 High-dimensionality of data
 Micro-array may have tens of thousands of dimensions
 High complexity of data
 Data streams and sensor data
 Time-series data, temporal data, sequence data
 Structure data, graphs, social networks and multi-linked data
 Heterogeneous databases and legacy databases
 Spatial, spatiotemporal, multimedia, text and Web data
 Software programs, scientific simulations
 New and sophisticated applications

29
 Applications of Data Mining
 Web page analysis: from web page classification, clustering to
PageRank & HITS algorithms
 Collaborative analysis & recommender systems
 Basket data analysis to targeted marketing
 Biological and medical data analysis: classification, cluster
analysis (microarray data analysis), biological sequence
analysis, biological network analysis
 Data mining and software engineering (e.g., IEEE Computer,
Aug. 2009 issue)
 From major dedicated data mining systems/tools (e.g., SAS, MS
SQL-Server Analysis Manager, Oracle Data Mining Tools) to
invisible data mining

31
 Major Issues in Data Mining (1)

 Mining Methodology
 Mining various and new kinds of knowledge
 Mining knowledge in multi-dimensional space
 Data mining: An interdisciplinary effort
 Boosting the power of discovery in a networked environment
 Handling noise, uncertainty, and incompleteness of data
 Pattern evaluation and pattern- or constraint-guided mining
 User Interaction
 Interactive mining
 Incorporation of background knowledge
 Presentation and visualization of data mining results

33
 Major Issues in Data Mining (2)

 Efficiency and Scalability

 Efficiency and scalability of data mining algorithms
 Parallel, distributed, stream, and incremental mining
methods
 Diversity of data types
 Handling complex types of data
 Mining dynamic, networked, and global data repositories
 Data mining and society
 Social impacts of data mining
 Privacy-preserving data mining
 Invisible data mining

34
 A Brief History of Data Mining Society

 1989 IJCAI Workshop on Knowledge Discovery in Databases

 Knowledge Discovery in Databases (G. Piatetsky-Shapiro and W.
Frawley, 1991)
 1991-1994 Workshops on Knowledge Discovery in Databases
 Advances in Knowledge Discovery and Data Mining (U. Fayyad, G.
Piatetsky-Shapiro, P. Smyth, and R. Uthurusamy, 1996)
 1995-1998 International Conferences on Knowledge Discovery in
Databases and Data Mining (KDD’95-98)
 Journal of Data Mining and Knowledge Discovery (1997)
 ACM SIGKDD conferences since 1998 and SIGKDD Explorations
 More conferences on data mining
 PAKDD (1997), PKDD (1997), SIAM-Data Mining (2001), (IEEE)
ICDM (2001), etc.
 ACM Transactions on KDD starting in 2007

36
 Conferences and Journals on Data Mining

 KDD Conferences  Other related conferences

 ACM SIGKDD Int. Conf. on
 DB conferences: ACM
Knowledge Discovery in
SIGMOD, VLDB, ICDE, EDBT,
Databases and Data Mining
(KDD) ICDT, …
 SIAM Data Mining Conf. (SDM)  Web and IR conferences:
 (IEEE) Int. Conf. on Data Mining WWW, SIGIR, WSDM
(ICDM)  ML conferences: ICML, NIPS
 European Conf. on Machine  PR conferences: CVPR,
Learning and Principles and  Journals
practices of Knowledge
Discovery and Data Mining
 Data Mining and Knowledge
(ECML-PKDD) Discovery (DAMI or DMKD)
 Pacific-Asia Conf. on Knowledge  IEEE Trans. On Knowledge and
Discovery and Data Mining Data Eng. (TKDE)
(PAKDD)  KDD Explorations
 Int. Conf. on Web Search and
 ACM Trans. on KDD
Data Mining (WSDM)
37
Where to Find References? DBLP, CiteSeer, Google

 Data mining and KDD (SIGKDD: CDROM)

 Conferences: ACM-SIGKDD, IEEE-ICDM, SIAM-DM, PKDD, PAKDD, etc.
 Journal: Data Mining and Knowledge Discovery, KDD Explorations, ACM TKDD
 Database systems (SIGMOD: ACM SIGMOD Anthology—CD ROM)
 Conferences: ACM-SIGMOD, ACM-PODS, VLDB, IEEE-ICDE, EDBT, ICDT, DASFAA
 Journals: IEEE-TKDE, ACM-TODS/TOIS, JIIS, J. ACM, VLDB J., Info. Sys., etc.
 AI & Machine Learning
 Conferences: Machine learning (ML), AAAI, IJCAI, COLT (Learning Theory), CVPR, NIPS,
etc.
 Journals: Machine Learning, Artificial Intelligence, Knowledge and Information Systems,
IEEE-PAMI, etc.
 Web and IR
 Conferences: SIGIR, WWW, CIKM, etc.
 Journals: WWW: Internet and Web Information Systems,
 Statistics
 Conferences: Joint Stat. Meeting, etc.
 Journals: Annals of statistics, etc.
 Visualization
 Conference proceedings: CHI, ACM-SIGGraph, etc.
 Journals: IEEE Trans. visualization and computer graphics, etc.

38
 Summary

 Data mining: Discovering interesting patterns and knowledge

from massive amount of data
 A natural evolution of database technology, in great demand,
with wide applications
 A KDD process includes data cleaning, data integration, data
selection, transformation, data mining, pattern evaluation, and
knowledge presentation
 Mining can be performed in a variety of data
 Data mining functionalities: characterization, discrimination,
association, classification, clustering, outlier and trend
analysis, etc.
 Data mining technologies and applications
 Major issues in data mining
40
 Recommended Reference Books
 S. Chakrabarti. Mining the Web: Statistical Analysis of Hypertex and Semi-Structured Data. Morgan Kaufmann,
2002
 R. O. Duda, P. E. Hart, and D. G. Stork, Pattern Classification, 2ed., Wiley-Interscience, 2000
 T. Dasu and T. Johnson. Exploratory Data Mining and Data Cleaning. John Wiley & Sons, 2003
 U. M. Fayyad, G. Piatetsky-Shapiro, P. Smyth, and R. Uthurusamy. Advances in Knowledge Discovery and Data
Mining. AAAI/MIT Press, 1996
 U. Fayyad, G. Grinstein, and A. Wierse, Information Visualization in Data Mining and Knowledge Discovery,
Morgan Kaufmann, 2001
 J. Han and M. Kamber. Data Mining: Concepts and Techniques. Morgan Kaufmann, 3 rd ed., 2011
 D. J. Hand, H. Mannila, and P. Smyth, Principles of Data Mining, MIT Press, 2001
 T. Hastie, R. Tibshirani, and J. Friedman, The Elements of Statistical Learning: Data Mining, Inference, and
Prediction, 2nd ed., Springer-Verlag, 2009
 B. Liu, Web Data Mining, Springer 2006.
 T. M. Mitchell, Machine Learning, McGraw Hill, 1997
 G. Piatetsky-Shapiro and W. J. Frawley. Knowledge Discovery in Databases. AAAI/MIT Press, 1991
 P.-N. Tan, M. Steinbach and V. Kumar, Introduction to Data Mining, Wiley, 2005
 S. M. Weiss and N. Indurkhya, Predictive Data Mining, Morgan Kaufmann, 1998
 I. H. Witten and E. Frank, Data Mining: Practical Machine Learning Tools and Techniques with Java
Implementations, Morgan Kaufmann, 2nd ed. 2005

41