DSS Chapter 1 - (Walid) Ben Ali
Download
Report
Transcript DSS Chapter 1 - (Walid) Ben Ali
Decision Support and Business
Intelligence Systems
(9th Ed., Prentice Hall)
Chapter 7:
Text and Web Mining and
text analytics
Learning Objectives
7-2
Describe text mining and understand the
need for text mining
Differentiate between text mining, Web
mining and data mining
Understand the different application areas for
text mining
Know the process of carrying out a text
mining project
Understand the different methods to
introduce structure to text-based data
Copyright © 2011 Pearson Education, Inc. Publishing as Prentice Hall
Learning Objectives
Describe Web mining, its objectives, and its
benefits
Understand the three different branches of
Web mining
7-3
Web content mining
Web structure mining
Web usage mining
Understand the applications of these three
mining paradigms
Copyright © 2011 Pearson Education, Inc. Publishing as Prentice Hall
7.1. Opening Vignette:
7-4
Copyright © 2011 Pearson Education, Inc. Publishing as Prentice Hall
Opening Vignette:
“Mining Text for Security and
Counterterrorism”
What is MITRE?
Problem description
Proposed solution
Results
Answer and discuss the case questions
7-5
Copyright © 2011 Pearson Education, Inc. Publishing as Prentice Hall
Opening Vignette:
Mining Text For Security…
Cluster 1
Cluster 3
(L) Kampala
Cluster 2
(P) Timothy McVeigh
(L) Uganda
(P) Oklahoma City
(P) Norodom Ranariddh
(P) Yoweri Museveni
(P) Terry Nichols
(P) Norodom Sihanouk
(E) election
(L) Sudan
(L) Bangkok
(L) Khartoum
(L) Cambodia
(L) Southern Sudan
(L) Phnom Penh
(L) Thailand
(P) Hun Sen
(O) Khmer Rouge
(P) Pol Pot
7-6
Copyright © 2011 Pearson Education, Inc. Publishing as Prentice Hall
7.2. Text Mining Concepts and
definitions
85-90 percent of all corporate data is in some
kind of unstructured form (e.g., text)
Unstructured corporate data is doubling in
size every 18 months
Tapping into these information sources is not
an option, but a need to stay competitive
Answer: text mining
7-7
A semi-automated process of extracting
knowledge from unstructured data sources
a.k.a. text data mining or knowledge discovery in
textual databases
Copyright © 2011 Pearson Education, Inc. Publishing as Prentice Hall
Data Mining versus Text Mining
Both seek for novel and useful patterns
Both are semi-automated processes
Difference is the nature of the data:
Structured versus unstructured data
7-8
Structured data: in databases
Unstructured data: Word documents, PDF files,
text excerpts, XML files, and so on
Text mining – first, impose structure to
the data, then mine the structured data
Copyright © 2011 Pearson Education, Inc. Publishing as Prentice Hall
Text Mining Concepts
Benefits of text mining are obvious especially
in text-rich data environments
Electronic communization records (e.g., Email)
7-9
e.g., law (court orders), academic research
(research articles), finance (quarterly reports),
medicine (discharge summaries), biology (molecular
interactions), technology (patent files), marketing
(customer comments), etc.
Spam filtering
Email prioritization and categorization
Automatic response generation
Copyright © 2011 Pearson Education, Inc. Publishing as Prentice Hall
Text Mining Application Area
7-10
Information extraction
Topic tracking
Summarization
Categorization
Clustering
Concept linking
Question answering
Copyright © 2011 Pearson Education, Inc. Publishing as Prentice Hall
Text Mining Terminology
7-11
Unstructured or semistructured data
Corpus (and corpora)
Terms
Concepts
Stemming
Stop words (and include words)
Synonyms (and polysemes)
Tokenizing
Copyright © 2011 Pearson Education, Inc. Publishing as Prentice Hall
Text Mining Terminology (Cont.)
Term dictionary
Word frequency
Part-of-speech tagging
Morphology
Term-by-document matrix
Singular value decomposition
7-12
Occurrence matrix
Latent semantic indexing
Copyright © 2011 Pearson Education, Inc. Publishing as Prentice Hall
Text Mining for Patent Analysis
(see Applications Case 7.2)
What is a patent?
How do we do patent analysis (PA)?
Why do we need to do PA?
7-13
“exclusive rights granted by a country to
an inventor for a limited period of time in
exchange for a disclosure of an invention”
What are the benefits?
What are the challenges?
How does text mining help in PA?
Copyright © 2011 Pearson Education, Inc. Publishing as Prentice Hall
Natural Language Processing (NLP)
Structuring a collection of text
NLP is …
7-14
Old approach: bag-of-words
New approach: natural language processing
a very important concept in text mining
a subfield of artificial intelligence and computational
linguistics
the studies of "understanding" the natural human
language
Syntax versus semantics based text mining
Copyright © 2011 Pearson Education, Inc. Publishing as Prentice Hall
Natural Language Processing (NLP)
What is “Understanding” ?
7-15
Human understands, what about computers?
Natural language is vague, context driven
True understanding requires extensive knowledge
of a topic
Can/will computers ever understand natural
language the same/accurate way we do?
Copyright © 2011 Pearson Education, Inc. Publishing as Prentice Hall
Natural Language Processing (NLP)
Challenges in NLP
Dream of AI community
7-16
Part-of-speech tagging
Text segmentation
Word sense disambiguation
Syntax ambiguity
Imperfect or irregular input
Speech acts
to have algorithms that are capable of automatically
reading and obtaining knowledge from text
Copyright © 2011 Pearson Education, Inc. Publishing as Prentice Hall
Natural Language Processing (NLP)
WordNet
Sentiment Analysis
7-17
A laboriously hand-coded database of English
words, their definitions, sets of synonyms, and
various semantic relations between synonym sets
A major resource for NLP
Need automation to be completed
A technique used to detect favorable and
unfavorable opinions toward specific products and
services
See Application Case 7.3 for a CRM application
Copyright © 2011 Pearson Education, Inc. Publishing as Prentice Hall
NLP Task Categories
7-18
Information retrieval
Information extraction
Named-entity recognition
Question answering
Automatic summarization
Natural language generation and understanding
Machine translation
Foreign language reading and writing
Speech recognition
Text proofing
Optical character recognition
Copyright © 2011 Pearson Education, Inc. Publishing as Prentice Hall
Text Mining Applications
Marketing applications
Security applications
Literature-based gene identification (…)
Academic applications
7-19
ECHELON, OASIS
Deception detection (…)
Medicine and biology
Enables better CRM
Research stream analysis
Copyright © 2011 Pearson Education, Inc. Publishing as Prentice Hall
Text Mining Applications
Application Case 7.4: Mining for Lies
Deception detection
The study
7-20
A difficult problem
If detection is limited to only text, then the
problem is even more difficult
analyzed text based testimonies of person
of interests at military bases
used only text-based features (cues)
Copyright © 2011 Pearson Education, Inc. Publishing as Prentice Hall
Text Mining Applications
Application Case 7.4: Mining for Lies
Statements
Transcribed for
Processing
Statements Labeled as
Truthful or Deceptive
By Law Enforcement
Cues Extracted &
Selected
Classification Models
Trained and Tested on
Quantified Cues
Text Processing
Software Identified
Cues in Statements
Text Processing
Software Generated
Quantified Cues
7-21
Copyright © 2011 Pearson Education, Inc. Publishing as Prentice Hall
Text Mining Applications
7-22
Application Case 7.4: Mining for Lies
Category
Example Cues
Quantity
Verb count, noun-phrase count, ...
Complexity
Avg. no of clauses, sentence length, …
Uncertainty
Modifiers, modal verbs, ...
Nonimmediacy
Passive voice, objectification, ...
Expressivity
Emotiveness
Diversity
Lexical diversity, redundancy, ...
Informality
Typographical error ratio
Specificity
Spatiotemporal, perceptual information …
Affect
Positive affect, negative affect, etc.
Copyright © 2011 Pearson Education, Inc. Publishing as Prentice Hall
Text Mining Applications
Application Case 7.4: Mining for Lies
371 usable statements are generated
31 features are used
Different feature selection methods used
10-fold cross validation is used
Results (overall % accuracy)
7-23
Logistic regression
Decision trees
Neural networks
67.28
71.60
73.46
Copyright © 2011 Pearson Education, Inc. Publishing as Prentice Hall
Text Mining Applications
Gene/
Protein
(gene/protein interaction identification)
596 12043 24224 281020
42722 397276
D007962
Ontology
D 016923
D 001773
D019254
D044465
D001769
D002477
D003643
D016158
Word
8 51112
9
23017
27
5874
2791
8952
1623
5632
17
8252
8 2523
NN
IN
NN
IN
VBZ
IN
JJ
JJ
NN
NN
NN
CC
NN
IN NN
NP
PP
NP
NP
PP NP
Shallow
Parse
185
POS
...expression of Bcl-2 is correlated with insufficient white blood cell death and activation of p53.
7-24
NP
PP
NP
Copyright © 2011 Pearson Education, Inc. Publishing as Prentice Hall
Text Mining Process
Context diagram for
the text mining
process
Unstructured data (text)
Structured data (databases)
Software/hardware limitations
Privacy issues
Linguistic limitations
Extract
Context-specific knowledge
knowledge
from available
data sources
A0
Domain expertise
Tools and techniques
7-25
Copyright © 2011 Pearson Education, Inc. Publishing as Prentice Hall
Text Mining Process
Task 1
Establish the Corpus:
Collect & Organize the
Domain Specific
Unstructured Data
Task 2
Create the TermDocument Matrix:
Introduce Structure
to the Corpus
Feedback
The inputs to the process
includes a variety of relevant
unstructured (and semistructured) data sources such
as text, XML, HTML, etc.
Task 3
The output of the Task 1 is a
collection of documents in
some digitized format for
computer processing
Extract Knowledge:
Discover Novel
Patterns from the
T-D Matrix
Feedback
The output of the Task 2 is a
flat file called term-document
matrix where the cells are
populated with the term
frequencies
The output of Task 3 is a
number of problem specific
classification, association,
clustering models and
visualizations
The three-step text mining process
7-26
Copyright © 2011 Pearson Education, Inc. Publishing as Prentice Hall
Text Mining Process
Step 1: Establish the corpus
7-27
Collect all relevant unstructured data
(e.g., textual documents, XML files, emails,
Web pages, short notes, voice recordings…)
Digitize, standardize the collection
(e.g., all in ASCII text files)
Place the collection in a common place
(e.g., in a flat file, or in a directory as
separate files)
Copyright © 2011 Pearson Education, Inc. Publishing as Prentice Hall
Text Mining Process
Step 2: Create the Term–by–Document Matrix
Terms
Documents
Document 1
in
ve
e
stm
ri
nt
jec
o
r
p
tm
a
an
ge
ftw
so
1
Document 2
are
t
g
en
in
v
de
elo
ng
e
pm
nt
SA
1
3
Document 4
1
1
Document 5
2
1
1
1
...
7-28
ri
ee
1
Document 3
Document 6
sk
n
me
Copyright © 2011 Pearson Education, Inc. Publishing as Prentice Hall
P
...
Text Mining Process
Step 2: Create the Term–by–Document
Matrix (TDM), cont.
Should all terms be included?
What is the best representation of the
indices (values in cells)?
7-29
Stop words, include words
Synonyms, homonyms
Stemming
Row counts; binary frequencies; log frequencies;
Inverse document frequency
Copyright © 2011 Pearson Education, Inc. Publishing as Prentice Hall
Text Mining Process
Step 2: Create the Term–by–Document
Matrix (TDM), cont.
TDM is a sparse matrix. How can we reduce
the dimensionality of the TDM?
7-30
Manual - a domain expert goes through it
Eliminate terms with very few occurrences in
very few documents (?)
Transform the matrix using singular value
decomposition (SVD)
SVD is similar to principle component analysis
Copyright © 2011 Pearson Education, Inc. Publishing as Prentice Hall
Text Mining Process
Step 3: Extract patterns/knowledge
Classification (text categorization)
Clustering (natural groupings of text)
7-31
Improve search recall
Improve search precision
Scatter/gather
Query-specific clustering
Association
Trend Analysis (…)
Copyright © 2011 Pearson Education, Inc. Publishing as Prentice Hall
Text Mining Application
(research trend identification in literature)
Mining the published IS literature
7-32
MIS Quarterly (MISQ)
Journal of MIS (JMIS)
Information Systems Research (ISR)
Covers 12-year period (1994-2005)
901 papers are included in the study
Only the paper abstracts are used
9 clusters are generated for further analysis
Copyright © 2011 Pearson Education, Inc. Publishing as Prentice Hall
Text Mining Application
(research trend identification in literature)
7-33
Journal Year
Author(s)
MISQ
2005
A. Malhotra,
S. Gosain and
O. A. El Sawy
ISR
1999
JMIS
2001
R. Aron and
E. K. Clemons
…
…
…
Title
Vol/No Pages
Absorptive capacity
configurations in
supply chains:
Gearing for partnerenabled market
knowledge creation
D. Robey and
Accounting for the
M. C. Boudreau contradictory
organizational
consequences of
information
technology:
Theoretical directions
and methodological
implications
Keywords
Abstract
145-187 knowledge management
supply chain
absorptive capacity
interorganizational
information systems
configuration approaches
2-Oct 167-185 organizational
transformation
impacts of technology
organization theory
research methodology
intraorganizational power
electronic communication
mis implementation
culture
systems
Achieving the optimal 18/2 65-88
information products
balance between
internet advertising
investment in quality
product positioning
and investment in selfsignaling
promotion for
signaling games
information products
…
29/1
…
…
…
Copyright © 2011 Pearson Education, Inc. Publishing as Prentice Hall
The need for continual value
innovation is driving supply
chains to evolve from a pure
transactional focus to
leveraging interorganizational
partner ships for sharing
Although much contemporary
thought considers advanced
information technologies as
either determinants or enablers
of radical organizational
change, empirical studies have
revealed inconsistent findings to
support the deterministic logic
implicit in such arguments. This
paper reviews the contradictory
When producers of goods (or
services) are confronted by a
situation in which their offerings
no longer perfectly match
consumer preferences, they
must determine the extent to
which the advertised features of
…
7-34
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
CLUSTER: 5
CLUSTER: 6
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
35
30
25
20
15
10
5
0
CLUSTER: 4
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
No of Articles
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
Text Mining Application
(research trend identification in literature)
35
30
25
20
15
10
5
0
35
30
25
20
15
10
5
0
CLUSTER: 1
CLUSTER: 2
CLUSTER: 3
CLUSTER: 7
CLUSTER: 8
CLUSTER: 9
YEAR
Copyright © 2011 Pearson Education, Inc. Publishing as Prentice Hall
Text Mining Application
(research trend identification in literature)
100
90
80
70
60
50
40
30
20
10
0
ISR
JMIS
MISQ
No of Articles
CLUSTER: 1
ISR
JMIS
MISQ
CLUSTER: 2
ISR
JMIS
MISQ
CLUSTER: 3
100
90
80
70
60
50
40
30
20
10
0
ISR
JMIS
MISQ
CLUSTER: 4
ISR
JMIS
MISQ
CLUSTER: 5
ISR
JMIS
MISQ
CLUSTER: 6
100
90
80
70
60
50
40
30
20
10
0
ISR
JMIS
MISQ
CLUSTER: 7
ISR
JMIS
MISQ
CLUSTER: 8
JOURNAL
7-35
Copyright © 2011 Pearson Education, Inc. Publishing as Prentice Hall
ISR
JMIS
MISQ
CLUSTER: 9
Text Mining (text analytics) Tools
Commercial Software Tools
Free Software Tools
7-36
SPSS PASW Text Miner
SAS Enterprise Miner
Statistica Data Miner
ClearForest, …
RapidMiner
GATE
Spy-EM, …
Copyright © 2011 Pearson Education, Inc. Publishing as Prentice Hall
Tools
7-37
Copyright © 2011 Pearson Education, Inc. Publishing as Prentice Hall
7-38
Copyright © 2011 Pearson Education, Inc. Publishing as Prentice Hall
ClearForest
7-39
Copyright © 2011 Pearson Education, Inc. Publishing as Prentice Hall
IBM Intelligent Miner, Data
Mining Suite
7-40
Copyright © 2011 Pearson Education, Inc. Publishing as Prentice Hall
Megaputer textanalyst
7-41
Copyright © 2011 Pearson Education, Inc. Publishing as Prentice Hall
SAS text miner
7-42
Copyright © 2011 Pearson Education, Inc. Publishing as Prentice Hall
Statistica text mining engine
7-43
Copyright © 2011 Pearson Education, Inc. Publishing as Prentice Hall
VantagePoint
7-44
Copyright © 2011 Pearson Education, Inc. Publishing as Prentice Hall
WordStat analysis module
7-45
Copyright © 2011 Pearson Education, Inc. Publishing as Prentice Hall
GATE
7-46
Copyright © 2011 Pearson Education, Inc. Publishing as Prentice Hall
LingPipe
7-47
Copyright © 2011 Pearson Education, Inc. Publishing as Prentice Hall
S-EM (Spy-EM)
7-48
Copyright © 2011 Pearson Education, Inc. Publishing as Prentice Hall
Vivisimo/Clusty
7-49
Copyright © 2011 Pearson Education, Inc. Publishing as Prentice Hall
Web Mining Overview
Web is the largest repository of data
Data is in HTML, XML, text format
Challenges (of processing Web data)
7-50
The
The
The
The
The
Web
Web
Web
Web
Web
is too big for effective data mining
is too complex
is too dynamic
is not specific to a domain
has everything
Opportunities and challenges are great!
Copyright © 2011 Pearson Education, Inc. Publishing as Prentice Hall
Web Mining
Web mining (or Web data mining) is the
process of discovering intrinsic relationships
from Web data (textual, linkage, or usage)
Web Mining
Web Content Mining
Source: unstructured
textual content of the
Web pages (usually in
HTML format)
7-51
Web Structure Mining
Source: the unified
resource locator (URL)
links contained in the
Web pages
Copyright © 2011 Pearson Education, Inc. Publishing as Prentice Hall
Web Usage Mining
Source: the detailed
description of a Web
site’s visits (sequence
of clicks by sessions)
Web Content/Structure Mining
Mining of the textual content on the Web
Data collection via Web crawlers
Web pages include hyperlinks
7-52
Authoritative pages
Hubs
hyperlink-induced topic search (HITS) alg
Copyright © 2011 Pearson Education, Inc. Publishing as Prentice Hall
Web Usage Mining
Extraction of information from data generated
through Web page visits and transactions…
7-53
data stored in server access logs, referrer logs,
agent logs, and client-side cookies
user characteristics and usage profiles
metadata, such as page attributes, content
attributes, and usage data
Clickstream data
Clickstream analysis
Copyright © 2011 Pearson Education, Inc. Publishing as Prentice Hall
Web Usage Mining
Web usage mining applications
7-54
Determine the lifetime value of clients
Design cross-marketing strategies across products.
Evaluate promotional campaigns
Target electronic ads and coupons at user groups
based on user access patterns
Predict user behavior based on previously learned
rules and users' profiles
Present dynamic information to users based on
their interests and profiles…
Copyright © 2011 Pearson Education, Inc. Publishing as Prentice Hall
Web Usage Mining
(clickstream analysis)
Pre-Process Data
Collecting
Merging
Cleaning
Structuring
- Identify users
- Identify sessions
- Identify page views
- Identify visits
Website
User /
Customer
Weblogs
How to better the data
How to improve the Web site
How to increase the customer value
7-55
Copyright © 2011 Pearson Education, Inc. Publishing as Prentice Hall
Extract Knowledge
Usage patterns
User profiles
Page profiles
Visit profiles
Customer value
Web Mining Success Stories
Amazon.com, Ask.com, Scholastic.com, …
Website Optimization Ecosystem
Customer Interaction
on the Web
Analysis of Interactions
Web
Analytics
Voice of
Customer
Customer Experience
Management
7-56
Copyright © 2011 Pearson Education, Inc. Publishing as Prentice Hall
Knowledge about the Holistic
View of the Customer
Web Mining Tools
7-57
Product Name
URL
Angoss Knowledge WebMiner
angoss.com
ClickTracks
clicktracks.com
LiveStats from DeepMetrix
deepmetrix.com
Megaputer WebAnalyst
megaputer.com
MicroStrategy Web Traffic Analysis
microstrategy.com
SAS Web Analytics
sas.com
SPSS Web Mining for Clementine
spss.com
WebTrends
webtrends.com
XML Miner
scientio.com
Copyright © 2011 Pearson Education, Inc. Publishing as Prentice Hall
Tools
7-58
Copyright © 2011 Pearson Education, Inc. Publishing as Prentice Hall
Angoss Knowledge WebMiner
7-59
Copyright © 2011 Pearson Education, Inc. Publishing as Prentice Hall
ClickTracks
7-60
Copyright © 2011 Pearson Education, Inc. Publishing as Prentice Hall
LiveStats from DeepMetrix
7-61
Copyright © 2011 Pearson Education, Inc. Publishing as Prentice Hall
Megaputer WebAnalyst
7-62
Copyright © 2011 Pearson Education, Inc. Publishing as Prentice Hall
MicroStrategy Web Traffic
Module
7-63
Copyright © 2011 Pearson Education, Inc. Publishing as Prentice Hall
Analysis
SAS Web Analytics
7-64
Copyright © 2011 Pearson Education, Inc. Publishing as Prentice Hall
SPSS Web Mining for Clementine
7-65
Copyright © 2011 Pearson Education, Inc. Publishing as Prentice Hall
WebTrends
7-66
Copyright © 2011 Pearson Education, Inc. Publishing as Prentice Hall
XML Miner
7-67
Copyright © 2011 Pearson Education, Inc. Publishing as Prentice Hall
7-68
Copyright © 2011 Pearson Education, Inc. Publishing as Prentice Hall
End of the Chapter
7-69
Questions / comments…
Copyright © 2011 Pearson Education, Inc. Publishing as Prentice Hall