What is Data Mining?

DAMA-NCR

•Tuesday, November 13, 2001

•Laura Squier
•Technical Consultant
•[email protected]
Agenda
• What Data Mining IS and IS NOT
• Steps in the Data Mining Process
– CRISP-DM
– Explanation of Models
– Examples of Data Mining
Applications
• Questions
 The Evolution of Data Analysis
Evolutionary Step Business Question Enabling Product Providers Characteristics
Technologies

Data Collection "What was my total Computers, tapes, IBM, CDC Retrospective,
(1960s) revenue in the last disks static data delivery
five years?"

Data Access "What were unit Relational Oracle, Sybase, Retrospective,

(1980s) sales in New databases Informix, IBM, dynamic data
England last (RDBMS), Microsoft delivery at record
March?" Structured Query level
Language (SQL),
ODBC

Data Warehousing "What were unit On-line analytic SPSS, Comshare, Retrospective,
& Decision sales in New processing Arbor, Cognos, dynamic data
Support England last (OLAP), Microstrategy,NCR delivery at multiple
(1990s) March? Drill down multidimensional levels
to Boston." databases, data
warehouses

Data Mining "What’s likely to Advanced SPSS/Clementine, Prospective,

(Emerging Today) happen to Boston algorithms, Lockheed, IBM, proactive
unit sales next multiprocessor SGI, SAS, NCR, information
month? Why?" computers, massive Oracle, numerous delivery
databases startups
Results of Data Mining
Include:
• Forecasting what may happen in
the future
• Classifying people or things into
groups by recognizing patterns
• Clustering people or things into
groups based on their attributes
• Associating what events are likely
to occur together
• Sequencing what events are likely
to lead to later events
 Data mining is not
•Brute-force crunching of bulk
data
•“Blind” application of algorithms
•Going to find relationships
where none exist
•Presenting data in different
ways
•A database intensive task
•A difficult to understand
technology requiring an
advanced degree in computer
science
Data Mining Is
•A hot buzzword for a class of
techniques that find patterns in data
•A user-centric, interactive process
which leverages analysis
technologies and computing power
•A group of techniques that find
relationships that have not
previously been discovered
•Not reliant on an existing database
•A relatively easy task that requires
knowledge of the business
problem/subject matter expertise
 Data Mining versus
OLAP
•OLAP - On-line
Analytical
Processing
– Provides you
with a very
good view of
what is
happening, but
can not predict
what will
happen in the
future or why it
is happening
 Data Mining Versus Statistical
Analysis
•Data Mining •Data Analysis
– Originally developed to act as – Tests for statistical correctness of
expert systems to solve problems models
– Less interested in the mechanics of • Are statistical assumptions of
the technique models correct?
– If it makes sense then let’s use it – Eg Is the R-Square good?
– Does not require assumptions to be – Hypothesis testing
made about data
• Is the relationship significant?
– Can find patterns in very large – Use a t-test to validate
amounts of data
significance
– Requires understanding of data and
business problem – Tends to rely on sampling
– Techniques are not optimised for
large amounts of data
– Requires strong statistical skills
 Examples of What People
are Doing with Data Mining:
•Fraud/Non-Compliance •Recruiting/Attracting
Anomaly detection customers
– Isolate the factors that lead to •Maximizing
fraud, waste and abuse profitability (cross
– Target auditing and
selling, identifying
profitable customers)
investigative efforts more
effectively •Service Delivery and
Customer Retention
•Credit/Risk Scoring
– Build profiles of
•Intrusion detection customers likely to
•Parts failure prediction use which services
•Web Mining
 How Can We Do Data
Mining?
By Utilizing the CRISP-
DM Methodology
– a standard process
– existing data
– software
technologies
– situational expertise
Why Should There be a
Standard Process?
•Framework for recording
experience
– Allows projects to be
The data mining process must replicated
be reliable and repeatable by •Aid to project planning and
people with little data mining management
•“Comfort factor” for new
background. adopters
– Demonstrates maturity of
Data Mining
– Reduces dependency on
“stars”
 Process
Standardization
CRISP-DM:
• CRoss Industry Standard Process for Data Mining
• Initiative launched Sept.1996
• SPSS/ISL, NCR, Daimler-Benz, OHRA
• Funding from European commission
• Over 200 members of the CRISP-DM SIG worldwide
– DM Vendors - SPSS, NCR, IBM, SAS, SGI, Data Distilleries,
Syllogic, Magnify, ..
– System Suppliers / consultants - Cap Gemini, ICL Retail, Deloitte
& Touche, …
– End Users - BT, ABB, Lloyds Bank, AirTouch, Experian, ...
CRISP-DM
•Non-proprietary
•Application/Industry
neutral
•Tool neutral
•Focus on business issues
– As well as technical
analysis
•Framework for guidance
•Experience base
– Templates for
Analysis
The
CRISP-
DM
Process
Model
 Why CRISP-DM?
•The data mining process must be reliable and repeatable by
people with little data mining skills

•CRISP-DM provides a uniform framework for

–guidelines
–experience documentation

•CRISP-DM is flexible to account for differences

–Different business/agency problems
–Different data
 Phases and Tasks
Business Data Data
Modeling Evaluation Deployment
Understanding Understanding Preparation

Determine Collect Initial Data Data Set Select Modeling Evaluate Results Plan Deployment
Business Objectives
Initial Data Collection Data Set Description Technique Assessment of Data Deployment Plan
Background Report Modeling Technique Mining Results w.r.t.
Business Objectives Select Data Modeling Assumptions Business Success Plan Monitoring and
Business Success Describe Data Rationale for Inclusion
/ Criteria Maintenance
Criteria Data Description Report Exclusion Generate Test DesignApproved Models Monitoring and
Test Design Maintenance Plan
Situation Assessment Explore Data Clean Data Review Process
Inventory of ResourcesData Exploration ReportData Cleaning Report Build Model Review of Process Produce Final Report
Requirements, Parameter Settings Final Report
Assumptions, and Verify Data Quality Construct Data Models Determine Next StepsFinal Presentation
Constraints Data Quality Report Derived Attributes Model Description List of Possible Actions
Risks and Contingencies Generated Records Decision Review Project
Terminology Assess Model Experience
Costs and Benefits Integrate Data Model Assessment Documentation
Merged Data Revised Parameter
Determine Settings
Data Mining Goal Format Data
Data Mining Goals Reformatted Data
Data Mining Success
Criteria

Produce Project Plan

Project Plan
Initial Asessment of
Tools and Techniques
Phases in the DM Process:
CRISP-DM
 Phases in the DM
Process (1 & 2)
•Business Understanding:
– Statement of
Business Objective
– Statement of Data •Data Understanding
Mining objective
– Explore the data and
– Statement of Success
verify the quality
Criteria
– Find outliers
 Phases in the DM
Process (3)
• Data preparation:
– Takes usually over 90% of our time
• Collection
• Assessment
• Consolidation and Cleaning
– table links, aggregation level, missing
values, etc
• Data selection
– active role in ignoring non-contributory
data?
– outliers?
– Use of samples
– visualization tools
• Transformations - create new variables
Phases in the DM Process
(4)
• Model building
– Selection of the modeling
techniques is based upon the data
mining objective
– Modeling is an iterative process -
different for supervised and
unsupervised learning
• May model for either description
or prediction
 Types of Models
•Prediction Models for •Descriptive Models for
Predicting and Classifying Grouping and Finding
– Regression algorithms Associations
(predict numeric outcome): – Clustering/Grouping
neural networks, rule algorithms: K-means,
induction, CART (OLS Kohonen
regression, GLM)
– Association algorithms:
– Classification algorithm apriori, GRI
predict symbolic outcome):
CHAID, C5.0 (discriminant
analysis, logistic regression)
Neural Network
Input layer
Hidden layer

Output
 Neural Networks
• Description
– Difficult interpretation
– Tends to ‘overfit’ the data
– Extensive amount of training time
– A lot of data preparation
– Works with all data types
 Rule Induction
•Description
– Produces decision trees:
• income < $40K
– job > 5 yrs then good risk
– job < 5 yrs then bad risk
• income > $40K C re d it ra n kin g (1 = d e fa u lt)

C a t. % n

– high debt then bad risk Ba d 5 2 .0 1 1 6 8

G o o d 4 7 .9 9 1 5 5
T o ta (l 1 0 0 .0 03)2 3

– low debt then good risk P a id W e e kly/Mo n th ly

P -va lu e =0 .0 0 0 0 , C h i-s q u a re =1 7 9 .6 6 6 5 , d f=1

– Or Rule Sets:
W e e kly p a y M o n th ly s a la ry

C a t. % n C a t. % n
B a d 8 6 .6 7 1 4 3 B a d 1 5 .8 2 2 5

• Rule #1 for good risk:

Go o d 1 3 .3 3 2 2 Good 8 4 .1 8 1 3 3
To ta l(5 1 .0 8 1) 6 5 T o ta l(4 8 .9 2 1) 5 8

Ag e C a te g o rica l Ag e C a te g o ric a l
P -va lu e = 0 .0 0 0 0 , C h i-s q u a re =3 0 .1 1 1 3 , d f=1 P -va lu e =0 .0 0 0 0 , C h i-s q u a re =5 8 .7 2 5 5 , d f=1

– if income > $40K Yo u n g (< 2 5 );Mid d le (2 5 -3 5 ) O ld ( > 3 5 ) Yo u n g (< 2 5 ) Mid d le (2 5 -3 5 );O ld ( > 3 5 )

– if low debt
C a t. % n C a t. % n C a t. % n C a t. % n
Ba d 9 0 .5 1 1 4 3 Ba d 0 .0 0 0 B a d 4 8 .9 8 2 4 B a d 0 .9 2 1
G o o d 9 .4 9 1 5 Good1 0 0 .0 0 7 Good 5 1 .0 2 2 5 Good 9 9 .0 8 1 0 8
To ta l(4 8 .9 2 1) 5 8 To ta l (2 .1 7 ) 7 To ta l(1 5 .1 7 )4 9 To ta l(3 3 .7 5 1) 0 9

• Rule #2 for good risk: S o c ia l C la s s

P -va lu e = 0 .0 0 1 6 , C h i-s q u a re = 1 2 .0 3 8 8 , d f=1

– if income < $40K Ma n a g e m e n t;C le rica l

C a t. % n
P ro fe s s io n a l

C a t. % n
B a d 0 .0 0 0 Ba d 5 8 .5 4 2 4

– if job > 5 years Good1 0 0 .0 0 8

To ta l (2 .4 8 ) 8
Go o d 4 1 .4 6 1 7
To ta l(1 2 .6 9 )4 1
 Rule Induction
Description
• Intuitive output
• Handles all forms of numeric data, as well
as non-numeric (symbolic) data

C5 Algorithm a special case of rule

induction
• Target variable must be symbolic
 Apriori
Description
• Seeks association rules in
dataset
• ‘Market basket’ analysis
• Sequence discovery
 Kohonen Network
Description
• unsupervised
• seeks to
describe
dataset in terms
of natural
clusters of
cases
 Phases in the DM
Process (5)
• Model Evaluation
– Evaluation of model: how well it
performed on test data
– Methods and criteria depend on
model type:
• e.g., coincidence matrix with
classification models, mean
error rate with regression
models
– Interpretation of model:
important or not, easy or hard
depends on algorithm
 Phases in the DM
Process (6)
•Deployment
– Determine how the results need to be
utilized
– Who needs to use them?
– How often do they need to be used
•Deploy Data Mining results by:
– Scoring a database
– Utilizing results as business rules
– interactive scoring on-line
Specific Data Mining
Applications:
What data mining has
done for...
The US Internal Revenue Service
needed to improve customer
service and...

Scheduled its workforce

to provide faster, more accurate
answers to questions.
What data mining has done
for...
The US Drug Enforcement
Agency needed to be more
effective in their drug “busts”
and

analyzed suspects’ cell phone

usage to focus investigations.
What data mining has done
for...
HSBC need to cross-sell more
effectively by identifying profiles
that would be interested in higher
yielding investments and...

Reduced direct mail costs by 30%

while garnering 95% of the
campaign’s revenue.
Final Comments
• Data Mining can be utilized in any
organization that needs to find
patterns or relationships in their
data.
• By using the CRISP-DM
methodology, analysts can have a
reasonable level of assurance that
their Data Mining efforts will
render useful, repeatable, and
valid results.
Questions?