DM: RE: Evaluation mining tools

[Alex Büchner]

>From: Harold van den Hauten [SMTP:hhauten@technologist.com]
>Sent: Wednesday, July 15, 1998 3:16 AM
>To: Datamine
>Subject: DM: Evaluation miningtools
>
>Does anyone have a pointer to a checklist for evauating datamining 
>tools?


Our latest book 'Decision Support using Data Mining' by Sarab S. Anand
and Alex B|chner has just been published by Financial Times Pitman
Publishers
(ISBN 0-273-63269-8). One appendix contains a comprehensive check 
list as
well as a market overview.

The table of contents is as following:

1 INTRODUCTION 1
1.1 DECISION SUPPORT SYSTEMS 1
1.2 THE DEMAND FOR THE NEW TECHNOLOGY 4
1.3 DATA MINING: LOOKING BEYOND THE TIP OF THE ICEBERG 7
1.4 THE CONTRIBUTION TO YOUR BUSINESS 8
1.5 SOME REPORTED DATA MINING SUCCESS STORIES 9
1.6 AUDIENCE AND OUTLINE OF TREATISE 10

2 A HISTORICAL PERSPECTIVE OF DATA MANAGEMENT 12
2.1 PROBLEMS AND ADVANTAGES 12
2.2 FROM TRANSACTION PROCESSING TO DECISION SUPPORT 13
2.3 WHERE IS ALL THE DATA AT THE MINUTE? 14
2.3.1 CLIENT-SERVER DATABASE ENVIRONMENTS 14
2.3.2 DATA WAREHOUSING TECHNOLOGY - LAYING THE IT FOUNDATION FOR 
DECISION
SUPPORT 15
2.3.3 MULTI-DIMENSIONAL DATABASES 17
2.4 THE FUTURE: WHAT CAN WE DO WITH ALL THAT DATA? 17

3 RELATED DISCIPLINES 19
3.1 DATABASE REPORTING TOOLS 19
3.2 STATISTICAL DATA ANALYSIS 19
3.3 MACHINE LEARNING TOOLS 21
3.4 ON-LINE ANALYTICAL PROCESSING 21

4 THE DATA MINING PROCESS 24
4.1 HUMAN RESOURCE IDENTIFICATION 24
4.2 PROBLEM SPECIFICATION 25
4.2.1 TASK IDENTIFICATION 26
4.2.2 USER IDENTIFICATION 26
4.3 DATA PROSPECTING 27
4.4 DOMAIN KNOWLEDGE ELICITATION 29
4.4.1 HIERARCHICAL GENERALISATION TREES 29
4.4.2 ATTRIBUTE RELATIONSHIP RULES 30
4.4.3 ATTRIBUTE DEPENDENCIES 30
4.4.4 ENVIRONMENT-BASED CONSTRAINTS 31
4.5 METHODOLOGY IDENTIFICATION 32
4.6 DATA PRE-PROCESSING 33
4.6.1 NOISE AND RESIDUAL VARIATION 34
4.6.2 DISCOVERING OUTLIERS 34
4.6.3 MISSING VALUES 35
4.6.4 DATA DIMENSIONALITY REDUCTION 35
4.6.5 DATA CODING AND QUANTISATION 35
4.6.6 DATA REDUCTION 36
4.7 PATTERN DISCOVERY 36
4.8 KNOWLEDGE POST-PROCESSING 37
4.8.1 KNOWLEDGE FILTERING 37
4.8.2 KNOWLEDGE VALIDATION 39
4.9 KNOWLEDGE MAINTENANCE 40
4.10 REFINEMENT PROCESS 41

5 DATA MINING GOALS AND METHODOLOGIES 42
5.1 DATA MINING GOALS 42
5.1.1 DISCOVERY OF ASSOCIATIONS 42
5.1.2 CLASSIFICATION 44
5.1.3 SEQUENTIAL PATTERN DISCOVERY 45
5.1.4 DISCOVERY OF CHARACTERISTICS 46
5.1.5 CLUSTERING OR DATA SEGMENTATION 47
5.1.6 DEVIATION DETECTION 48
5.1.7 REGRESSION 49
5.1.8 TEMPORAL MODELLING 50
5.2 DATA MINING METHODOLOGIES 50
5.2.1 TRADITIONAL STATISTICS 51
5.2.2 MACHINE LEARNING TECHNIQUES 52
5.2.3 UNCERTAINTY-BASED TECHNIQUES 54
5.2.4 DATABASE TECHNIQUES 56
5.2.5 VISUALISATION 56
5.2.6 HYBRID TECHNIQUES 56

6 THE DATA MINING STRATEGY 57
6.1 COMMON MISTAKES, RISKS AND COSTS 57
6.1.1 COMMON MISTAKES AND RISKS 57
6.1.2 HIDDEN COSTS OF DATA MINING 58
6.2 SETTING UP A DATA MINING ENVIRONMENT 58
6.3 BUSINESS PROCESS REENGINEERING TO FACILITATE DATA MINING 59
6.3.1 OVERVIEW OF BPR CONCEPTS 59
6.3.2 THE BPR AND DATA MINING SYNTHESIS 63
6.3.3 ENTERPRISE REQUIREMENTS TO ALLOW DATA MINING 65

Appendix I: Case Studies 67
Appendix II: A Brief Market Overview 81
Appendix III: Glossary 88
Appendix IV: References 92
Appendix V: Further Reading 94
Appendix VI: Internet Resources 96

Check http://www.ftmanagement.com/ for more details.

:-) aleX
_________________________________________________
Alex G. B|chner, Research Fellow
Northern Ireland Knowledge Engineering Laboratory
University of Ulster, Newtownabbey, BT37 0QB,  UK
Phone:+44 (0)1232 368394  Fax: +44 (0)1232 366068
URL:  http://www.infj.ulst.ac.uk/staff/ag.buchner