1Machine Learning Methodsfor
Text / Web Data Mining
Byoung-Tak ZhangSchool of Computer Science and Engineering
Seoul National UniversityE-mail: [email protected]
This material is available at http://scai.snu.ac.kr./~btzhang/
2
Overview
z Introduction4Web Information Retrieval 4Machine Learning (ML)4ML Methods for Text/Web Data Mining
z Text/Web Data Analysis4Text Mining Using Helmholtz Machines4Web Mining Using Bayesian Networks
z Summary 4Current and Future Work
23
Web Information RetrievalText Data
Classification System
Information Filtering System
questionuser profile
feedback
answer
DB
LocationDate
DB Record
DB Template Filling& InformationExtraction System
Information FilteringInformation Extraction
filtered data
Text Classification
Preprocessing and Indexing
4
Machine Learning
z Supervised Learning4Estimate an unknown mapping from known input-
output pairs4Learn fw from training set D={(x,y)} s.t.
4Classification: y is discrete4Regression: y is continuous
z Unsupervised Learning4Only input values are provided4Learn fw from D={(x)} s.t.
4Density Estimation4Compression, Clustering
)()( xxw fyf ==
xxw =)(f
35
Machine Learning Methodsz Neural Networks
4Multilayer Perceptrons (MLPs)4Self-Organizing Maps (SOMs)4Support Vector Machines (SVMs)
z Probabilistic Models4Bayesian Networks (BNs)4Helmholtz Machines (HMs)4Latent Variable Models (LVMs)
z Other Machine Learning Methods4Evolutionary Algorithms (EAs)4Reinforcement Learning (RL)4Boosting Algorithms4Decision Trees (DTs)
6
ML for Text/Web Data Mining
z Bayesian Networks for Text Classificationz Helmholtz Machines for Text Clustering/Categorizationz Latent Variable Models for Topic Word Extractionz Boosted Learning for TREC Filtering Taskz Evolutionary Learning for Web Document Retrievalz Reinforcement Learning for Web Filtering Agentsz Bayesian Networks for Web Customer Data Mining
47
Preprocessing for Text LearningFrom: [email protected]:comp.graphicsSubject: Need specs on Apple QT
I need to the specs, or at least a very verbose interpretation of the specs, for QuckTime. Technical articles from magazines and references to books would be nice, too.
I also need the specs in a format usable on a Unix or MS-Dos system. I cant do much with the QuickTime stuff they have on..
specs3unix1
hockey0
computer0
.
space0references1
.
.quicktime2
graphics0
clinton0car0baseball0
8
Text Mining: Data Sets
z Usenet Newsgroup Data420 categories 41000 documents for each category 420000 documents in total.
z TDT2 Corpus4Target detection and tracking (TDT): NIST
4Used 6,169 documents in experiments
59
d1 d2 d3 dn
h1 h2 hm
4 [Chang and Zhang, 2000]4 Input nodes
Binary values Represent the existence or
absence of words in documents.
Text Mining:Helmholtz Machine Architecture
+
===
n
jjiji
i
dwbhP
1exp1
1)1(
+
===
m
jjiji
i
hwbdP
1exp1
1)1(
4Latent nodes Binary values Extract the underlying causal
structure in the document set. Capture correlations of the words
in documents.
: recognition weight
: generative weight
10
Text Mining: Learning Helmholtz Machines
4Introduce a recognition network for estimation of a generative network.
4Wake-Sleep Algorithm Train the recognition and generative models alternately. Update the weight in network iteratively by simple local delta rule.
( ) ( ) ( )( )( ) ( )( )
=
==
=
=T
tt
ttt
tt
ttt
T
t t
tt
t QdPQ
QdPQdPD
1)(
)()()(
T
1t )()(
)()()(
1 )(
)()(
)(
|,log
|,log |,log)|log(
))1(( ==+=
jjiij
ijoldij
newij
spsswwww
611
Text Mining: Methods
z Text Categorization4Train a Helmholtz machine for each category.4Total N machines for N categories.
4Once the N machines have been estimated, classification of a test document proceeds by estimating the likelihood of the document for each machine.
z Topic Words Extraction4For the entire document sets, train a Helmholtz machine.4After training, examine the weights of connections from a latent
node to input nodes, that is words.
)]|([logmaxarg cdPcCc
=
12
4Usenet Newsgroup Data 20 categories, 1000 documents for each category, 20000 documents in
total.
Text Mining: Categorization Results
713
4TDT2 Corpus46,169 documents
imf, monetary, currencies, currency, rupiah, singapore, bailout, traders, markets, thailand, inflation, investors, fund, banks, baht
7
netanyahu, palestinian, arafat, israeli, yasser, kofi, annan, benjamin, palestinians, mideast, gaza,jerusalem, eu, paris, israel
4
India, pakistan, pakistani, delhi, hindu, vajpayee, nuclear, tests, atal, kashmir, indian, janata, bharatiya, islamabad, bihari
5
Suharto, habibie, demonstrators, riots, indonesians, demonstrations, soeharto, resignation, jakarta, rioting, electoral, rallies, wiranto, unrest, megawati
6
pope, cuba, cuban, embargo, castro, lifting, cubans, havana, alan, invasion, reserve, paul, output, vatican, freedom
8
olympics, nagano, olympic, winter, medal, hockey, atheletes, cup, games, slalom, medals, bronze, skating, lillehammer, downhill
3
warplane, airline, saudi, gulf, wright, soldiers, yitzhak, tanks, stealth, sabah, stations, kurds, mordechai, separatist, governor
2
tabacco, smoking, gingrich, newt, trent, republicans, congressional, republicans, attorney, smokers, lawsuit, senate, cigarette, morris, nicotine
1
Text Mining: Topic Words Extraction Results
14
z KDD-2000 Web Mining Competition4Data: 465 features over 1700 customers
Features include friend promotion rate, date visited, weight of items, price of house, discount rate,
Data was collected during Jan. 30 March 30, 2000 Friend promotion was started from Feb. 29 with TV
advertisement. 4Aims: Description of heavy/low spenders
Web Mining: Customer Analysis
815
16
917
Web Mining: Feature Selectionz Features selected by various ways [Yang & Zhang, 2000]
V240 (Friend) V229 (Order-Average) V304 (OrderShippingAmtMin.)V368 (Weight Average)V43 (Home Market Value)V377 (NumAcountTemplate Views)+V11 (WhichDoYouWearMostFrequent)V13 (SendEmail)V17 (USState)V45 (VehicleLifeStyle)V68 (RetailActivity)V19 (Date)
V13 (SendEmail)V234 (OrderItemQuantity Sum%HavingDiscountRange(5 . 10))V237 (OrderItemQuantitySum%Having DiscountRange(10.))V240 (Friend)V243 (OrderLineQuantitySum)V245 (OrderLineQuantity Maximum)V304 (OrderShippingAmtMin)V324 (NumLegwearProductViews)V368 (Weight Average)V374 (NumMainTemplateViews)V412 (NumReplenishableStock Views)
V368 (Weight Average)V243 (OrderLine Quantity Sum)V245 (OrderLine Quantity Maximum)F1 = 0.94*V324 + 0.868*V374 + 0.898*V412 F2 = 0.829*V234 + 0.857*V240F3 = -0.795*V237+ 0.778*V304
Discriminant ModelDecision TreeDecisionTree+Factor Analysis
18
Web Mining: Bayesian Nets
z Bayesian network4DAG (Directed Acyclic Graph) 4Express dependence relations between variables 4Can use prior knowledge on the data (parameters)
A B C P(A,B,C,D,E) = P(A)P(B|A)P(C|B)
P(D|A,B)P(E|B,C,D)
D E
4Examples of conjugate priors:Dirichlet for multinomial data, Normal-Wishart for normal data
10
19
Web Mining: Results
z A Bayesian net for KDD web data
z V229 (Order-Average) and V240 (Friend) directly influence V312 (Target)
z V19 (Date) was influenced by V240 (Friend) reflecting the TV advertisement.
20
Summaryz We study machine learning methods, such as
4Probabilistic neural networks4Evolutionary algorithms4Reinforcement learning
z Application areas include4Text mining4Web mining4Bioinformatics (not addressed in this talk)
z Recent work focuses on probabilistic graphical models for web/text/bio data mining, including
4Bayesian networks4Helmholtz machines4Latent variable models
11
21
22
Bayesian Networks:Architecture
z A Bayesian network represents the probabilistic relationships between the variables.
B
G M
L
=
=n
iiiXPP
1
)|()( paX pai is the set of parent nodes of Xi.
),|()|()()( ),,|(),|()|()(),,,(
LBMPBGPBPLPGBLMPBLGPLBPLPMGBLP
==
12
23
z The network structure represents the nave Bayes assumption.z All nodes are binary.z [Hwang & Zhang, 2000]
C
t1 t2 t8754
C: document class
ti: ith term
Bayesian Networks:Applications in IR A Simple BN for Text Classification
24
z Dataset4The acq dataset from Reuters-2157848754 terms were selected by TFIDF.
4Training data: 8762 documents4Test data: 3009 documents
z Parametric Learning4Dirichlet prior assumptions for the network parameter
distributions.
4Parameter distributions are updated with training data.),...,|(Dir)|( 1 iijrijij
hij Sp =
),...,|(Dir),|( 11 ii ijrijrijijijh
ij NNSDp ++=
Bayesian Networks:Experimental Results
13
25
z For training data4Accuracy: 94.28%
z For test data4Accuracy: 96.51%
99.3293.76Negative examples75.9896.83Positive examples
Precision (%)Recall (%)
98.6796.88Negative examples89.1795.16Positive examples
Precision (%)Recall (%)
Bayesian Networks:Experimental Results
26
Latent Variable Models:Architecture
Latent Variable Model forTopic Words Extraction and Document Clustering
z [Shin & Zhang, 2000]z Maximize log-likelihood
4Update , , . 4With EM Algorithm
= = =
= =
=
=N
n
M
m
K
kknkmkmn
N
n
M
mmnmn
zdPzwPzPwdn
wdPwdnL
1 1 1
1 1
)|()|()(log),(
),(log),(
)( kzP )|( km zwP )|( kn zdP
Document Clustering
Topic-Words Extraction
14
27
z EM (Expectation-Maximization) Algorithm4Algorithm to maximize pre-defined log-likelihood
z Iteration of E-Step and M-Step4E-Step M-Step
=
= Kk
kmknk
kmknk
mnk
zwPzdPzP
zwPzdPzPwdzP
1)|()|()(
)|()|()(),|(
= =
== Mm
N
nmnkmn
N
nmnkmn
km
wdzPwdn
wdzPwdnzwP
1 1
1
),|(),(
),|(),()|(
= =
== Mm
N
nmnkmn
M
mmnkmn
kn
wdzPwdn
wdzPwdnzdP
1 1
1
),|(),(
),|(),()|(
= =
= =
=M
m
N
nmn
M
m
N
nmnkmnk
wdnR
wdzPwdnR
zP
1 1
1 1
),(
,),|(),(1)(
Latent Variable Models:Learning
28
z Topic Words Extraction and Document Clustering with a subset of TREC-8 data
z TREC-8 adhoc task data4Documents: DTDS, FR94, FT, FBIS, LATIMES4Topics: 401-450 (401, 434, 439, 450)4401: Foreign Minorities, Germany4434: Estonia, Economy4439: Inventions, Scientific discovery4450: King Hussein, Peace
Latent Variable Models:Applications in IR Experimental Results
15
29
0.9900.729290003450 (293)0.9270.953920307439 (219)0.6860.996791023820434 (347)0.9300.9022001279401 (300)RecallPrecisionz1z3z4z2Topic (#Docs)
Label (assigned to zk with Maximum P(di|zk) )
jordan, peac, isreal, palestinian, king, isra, arab, meet, talk, husayn, agreem, presid, majesti, negoti, minist, visit, region, arafat, secur, peopl, east, washington, econom, sign, relat, jerusalem, rabin, syria, iraq,
Cluster 1(z1)
research, technology, develop, mar, materi, system, nuclear, environment, electr, process, product, power, energi, countrol, japan, pollution, structur, chemic, plant,
Cluster 3(z3)
percent, estonia, bank, state, privat, russian, year, enterprise, trade, million, trade, estonian, econom, countri, govern, compani, foreign, baltic, polish, loan, invest, fund, product,
Cluster 4(z4)
german, germani, mr, parti, year, foreign, people, countri, govern, asylum, polit, nation, law, minist, europ, state, immigr, democrat, wing, social, turkish, west, east, member, attack,
Cluster 2(z2 )
Extracted Topic Words (top 35 words with highest P(wj|zk) Topics
Latent Variable Models:Applications in IR Experimental Results
30
Boosting: Algorithmsz A general method of converting rough rules into a highly accurate
prediction rulez Learning procedure
4 Examine the training set4 Derive a rough rule (weak learner)4 Re-weight the examples in the training set, concentrating on the hard cases for
previous rules4 Repeat T times
Learner Learner Learner Learner
h1 h2 h3 h4 ),,,( 4321 hhhhf
Importance weightsof training documents
16
31
Boosting: Applied to Text Filteringz Nave Bayes
4Traditional algorithm for text filtering
z Boosting nave Bayes4Using nave Bayes classifiers as weak learners4 [Kim & Zhang, SIGIR-2000]
)|""(
... )|""()|""()(maxarg
)|()(maxarg
)|()(maxarg
21
1
} ,{
jin
jijijc
n
kjikj
c
jijirrelevantrelevantc
NM
ctroublewP
capproachwPcourwPcP
cwPcP
cdPcPc
j
j
j
====
=
=
=
Assume independence among terms
32
z TREC (Text Retrieval Conference)4 Sponsored by NIST
z TREC-7 filtering datasets4Training Documents
AP articles (1988) 237 MB, 79919 documents4Test Documents
AP articles (1989~1990) 471 MB, 162999 documents4No. of topics: 50
Example of a document
z TREC-8 filtering datasets4Training Documents
Financial Times (1991~1992) 167 MB, 64139 documents4Test Documents
Financial Time (1993~1994) 382 MB, 140651 documents4No. of topics: 50
Boosting: Applied to Text Filtering Experimental Results
17
33
0.509
PIRC
0.500
PIRC
0.505
NTTATTBoostingNTTATTBoosting
0.4600.467
Averaged Scaled F3
0.4520.4610.474
Averaged Scaled F1
0.720
Mer
0.714
PIRC
0.734
PIRCCLBoostingPLT2PLT1Boosting
0.7210.722
Averaged Scaled LF2
0.7130.7120.717
Averaged Scaled LF1
TREC-7
TREC-8
Compared with the state-of-the-art text filtering systems
Boosting: Applied to Text Filtering Experimental Results
34
Evolutionary Learning:Applications in IR - Web-Document Retrieval
z [Kim & Zhang, 2000]
Link Information,HTML Tag Information
chromosomes
Retrieval
Fitness
wnw3w2w1
wnw3w2w1 wnw3w2w1wnw3w2w1
wnw3w2w1 wnw3w2w1
18
35
Evolutionary Learning:Applications in IR Tag Weighting
z Crossover
z Truncation selection
z Mutation
yny3y2y1xnx3x2x1
chromosome X chromosome Y
chromosome Z (offspring)
zi = (xi + yi ) / 2 w.p. Pc
chromosome X
chromosome X
change value w.p. Pm
xnx3x2x1
xnx3x2x1znz3z2z1
36
Evolutionary Learning :Applications in IR - Experimental Resultsz Datasets
4TREC-8 Web Track Data42GB, 247491 web documents (WT2g)4No. of training topics: 10, No. of test topics: 10
z Results
19
37
Agent
Environment
Reinforcement Learning:Basic Concept
1. State st Reward rt
3. Reward rt+1
4. State st+1
2. Action at
38
WAIR
User
2. Actioni(modify profile)
Rewardi
1. Statei(user profile)
User profile
...Filtered documents
Document filtering
4. Statei+1
3. Rewardi+1 (relevance feedback)
retrieve documents calculate similarity
Reinforcement Learning:Applications in IR - Information Filtering[Seo & Zhang, 2000]
20
39
(%)
Reinforcement Learning:Experimental Results (Explicit Feedback)
40
(%)
Reinforcement Learning:Experimental Results (Implicit Feedback)