龙星计划课程 : 信息检索 course overview & background
Post on 06-Jan-2016
141 Views
Preview:
DESCRIPTION
TRANSCRIPT
2008 © ChengXiang Zhai Dragon Star Lecture at Beijing University, June 21-30, 2008 1
龙星计划课程 :信息检索 Course Overview & Background
ChengXiang Zhai (翟成祥 ) Department of Computer Science
Graduate School of Library & Information Science
Institute for Genomic Biology, Statistics
University of Illinois, Urbana-Champaign
http://www-faculty.cs.uiuc.edu/~czhai, czhai@cs.uiuc.edu
2008 © ChengXiang Zhai Dragon Star Lecture at Beijing University, June 21-30, 2008 2
Outline
• Course overview
• Essential background
– Probability & statistics
– Basic concepts in information theory
– Natural language processing
2008 © ChengXiang Zhai Dragon Star Lecture at Beijing University, June 21-30, 2008 3
Course Overview
2008 © ChengXiang Zhai Dragon Star Lecture at Beijing University, June 21-30, 2008 4
Course Objectives
• Introduce the field of information retrieval (IR)
– Foundation: Basic concepts, principles, methods, etc
– Trends: Frontier topics
• Prepare students to do research in IR and/or related fields
– Research methodology (general and IR-specific)
– Research proposal writing
– Research project (to be finished after the lecture period)
2008 © ChengXiang Zhai Dragon Star Lecture at Beijing University, June 21-30, 2008 5
Prerequisites
• Proficiency in programming (C++ is needed for assignments)
• Knowledge of basic probability & statistics (would be necessary for understanding algorithms deeply)
• Big plus: knowledge of related areas
– Machine learning
– Natural language processing
– Data mining
– …
2008 © ChengXiang Zhai Dragon Star Lecture at Beijing University, June 21-30, 2008 6
Course Management
• Teaching staff– Instructor: ChengXiang Zhai (UIUC)
– Teaching assistants:
• Hongfei Yan (Peking Univ)
• Bo Peng (Peking Univ)
• Course website: http://net.pku.edu.cn/~course/cs410/
• Course group discussion: http://groups.google.com/group/cs410pku
• Questions: First post the questions on the group discussion forum; if questions are unanswered, bring them to the office hours (first office hour: June 23, 2:30-4:30pm)
2008 © ChengXiang Zhai Dragon Star Lecture at Beijing University, June 21-30, 2008 7
Format & Requirements
• Lecture-based:
– Morning lectures: Foundation & Trends
– Afternoon lectures: IR research methodology
– Readings are usually available online
• 2 Assignments (based on morning lectures)
– Coding (C++), experimenting with data, analyzing results, open explorations (~5 hours each)
• Final exam (based on morning lectures): 1:30-4:30pm, June 30.
– Practice questions will be available
2008 © ChengXiang Zhai Dragon Star Lecture at Beijing University, June 21-30, 2008 8
Format & Requirements (cont.)
• Course project (Mini-TREC)
– Work in teams
– Phase I: create test collections (~ 3 hours, done within lecture period)
– Phase II: develop algorithms and submit results (done in the summer)
• Research project proposal (based on afternoon lectures)
– Work in teams
– 2-page outline done within lecture period
– full proposal (5 pages) due later
2008 © ChengXiang Zhai Dragon Star Lecture at Beijing University, June 21-30, 2008 9
Coverage of Topics: IR vs. TIM
Text Information Management(TIM)
Information Retrieval(IR)
Multimedia, etc
IR and TIM will be used interchangeably
2008 © ChengXiang Zhai Dragon Star Lecture at Beijing University, June 21-30, 2008 10
What is Text Info. Management?
• TIM is concerned with technologies for managing and exploiting text information effectively and efficiently
• Importance of managing text information
– The most natural way of encoding knowledge• Think about scientific literature
– The most common type of information• How much textual information do you produce and
consume every day?
– The most basic form of information• It can be used to describe other media of
information
– The most useful form of information!
2008 © ChengXiang Zhai Dragon Star Lecture at Beijing University, June 21-30, 2008 11
Text Management Applications
Access Mining
Organization
Select information
Create Knowledge
Add Structure/Annotations
2008 © ChengXiang Zhai Dragon Star Lecture at Beijing University, June 21-30, 2008 12
Examples of Text Management Applications
• Search– Web search engines (Google, Yahoo, …)
– Library systems
– …
• Recommendation– News filter
– Literature/movie recommender
• Categorization– Automatically sorting emails
– …
• Mining/Extraction– Discovering major complaints from email in customer service
– Business intelligence
– Bioinformatics
– …
• Many others…
2008 © ChengXiang Zhai Dragon Star Lecture at Beijing University, June 21-30, 2008 13
Elements of Text Info Management Technologies
Search
Text
Filtering
Categorization
Summarization
Clustering
Natural Language Content Analysis
Extraction
Mining
VisualizationRetrievalApplications
MiningApplications
InformationAccess
KnowledgeAcquisition
InformationOrganization
Focus of the course
2008 © ChengXiang Zhai Dragon Star Lecture at Beijing University, June 21-30, 2008 14
Text Management and Other Areas
TM Algorithms
User
Text Storage
Compression
Probabilistic inferenceMachine learning
Natural language processing
Human-computer interaction
TM Applications Software engineeringWeb
Computer science
InformationScience
2008 © ChengXiang Zhai Dragon Star Lecture at Beijing University, June 21-30, 2008 15
Related Areas
InformationRetrieval Databases
Library & InfoScience
Machine LearningPattern Recognition
Data Mining
NaturalLanguageProcessing
ApplicationsWeb, Bioinformatics…
StatisticsOptimization
Software engineeringComputer systems
Models
Algorithms
Applications
Systems
2008 © ChengXiang Zhai Dragon Star Lecture at Beijing University, June 21-30, 2008 16
Publications/Societies (Incomplete)
ACM SIGIR
VLDB, PODS, ICDE
ASIS
Learning/Mining
NLP
Applications
Statistics
Software/systems
COLING, EMNLP, ANLP
HLT
ICML, NIPS, UAIRECOMB, PSB
JCDL
Info. Science
Info Retrieval
ACM CIKM
DatabasesACM SIGMOD
ACL
ICML
AAAI
ACM SIGKDD
ISMB WWW
SOSP
OSDI
TREC
2008 © ChengXiang Zhai Dragon Star Lecture at Beijing University, June 21-30, 2008 17
Schedule: available at http://net.pku.edu.cn/~course/cs410/
2008 © ChengXiang Zhai Dragon Star Lecture at Beijing University, June 21-30, 2008 18
Date Morning Lecture (8:30-11:30)(Foundation & Trends)
Afternoon Lecture (1:30-2:30) (Research Methodology)
Notes
6/21 Sat
Course overview and background (probability, statistics, information theory, NLP) Slides: pptLecture Notes: Prob & Stat, Info Theory, NLPReadings:Bush 45, Rosenfeld's note on estimation, Rosenfeld's note on information theory,
Introduction to IR research Slides: ppt
6/22 Sun
Information Retrieval Overview (part 1) (basic concepts, history, evaluation) Lecture Notes: text retrieval, Readings: Singhal's review (Error), Book-Ch8. TREC measures
Prepare yourself for IR research Mini-TREC task specification ready
6/23 Mon
Information Retrieval Overview (Part 2) (basic retrieval models, system implementation, applications)
Find a good IR research topic Assign #1 out
6/24 Tue
Statistical Language Models for IR (probabilistic retrieval models, KL-divergence model, special retrieval tasks)
Formulate IR research hypotheses
Assign #2 out
6/25 Wed
Modern Retrieval Frameworks (axiomatic, decision-theoretic)Final exam practice questions available
6/26 Thu
Personalized Search & User Modeling (implicit feedback, explicit feedback, active feedback)
Test/Refine IR research hypotheses
Proposal team due
6/27 Fri Natural Language Processing for IR (phrase indexing, dependency analysis, sense disambiguation, sentiment retrieval)
Write and publish an IR paper
6/28 Sat No class Mini-TREC Phase I Task due
6/29 Sun
Topic Models for Text mining (PLSA, LDA, extensions and applications)
Proposal outline due
6/30 Mon
Future of IR, course summary Final Exam (1:30-4:30) Assigns #1, #2 due
7/5 Sat Research proposal due
7/? Mini-TREC data sets ready
8/? Mini-TREC Phase II Task due
2008 © ChengXiang Zhai Dragon Star Lecture at Beijing University, June 21-30, 2008 19
Essential Backgroud 1:
Probability & Statistics
2008 © ChengXiang Zhai Dragon Star Lecture at Beijing University, June 21-30, 2008 20
Prob/Statistics & Text Management
• Probability & statistics provide a principled way to quantify the uncertainties associated with natural language
• Allow us to answer questions like:– Given that we observe “baseball” three times and “game” once
in a news article, how likely is it about “sports”? (text categorization, information retrieval)
– Given that a user is interested in sports news, how likely would the user use “baseball” in a query? (information retrieval)
2008 © ChengXiang Zhai Dragon Star Lecture at Beijing University, June 21-30, 2008 21
Basic Concepts in Probability
• Random experiment: an experiment with uncertain outcome (e.g., tossing a coin, picking a word from text)
• Sample space: all possible outcomes, e.g.,
– Tossing 2 fair coins, S ={HH, HT, TH, TT}
• Event: ES, E happens iff outcome is in E, e.g.,
– E={HH} (all heads)
– E={HH,TT} (same face)
– Impossible event ({}), certain event (S)
• Probability of Event : 1P(E) 0, s.t.
– P(S)=1 (outcome always in S)
– P(A B)=P(A)+P(B) if (AB)= (e.g., A=same face, B=different face)
2008 © ChengXiang Zhai Dragon Star Lecture at Beijing University, June 21-30, 2008 22
Basic Concepts of Prob. (cont.)
• Conditional Probability :P(B|A)=P(AB)/P(A)
– P(AB) = P(A)P(B|A) =P(B)P(A|B)
– So, P(A|B)=P(B|A)P(A)/P(B) (Bayes’ Rule)
– For independent events, P(AB) = P(A)P(B), so P(A|B)=P(A)
• Total probability: If A1, …, An form a partition of S, then
– P(B)= P(BS)=P(BA1)+…+P(B An) (why?)
– So, P(Ai|B)=P(B|Ai)P(Ai)/P(B)
= P(B|Ai)P(Ai)/[P(B|A1)P(A1)+…+P(B|An)P(An)]
– This allows us to compute P(Ai|B) based on P(B|Ai)
2008 © ChengXiang Zhai Dragon Star Lecture at Beijing University, June 21-30, 2008 23
Interpretation of Bayes’ Rule
)(
)()|()|(
EP
HPHEPEHP ii
i
Hypothesis space: H={H1 , …, Hn} Evidence: E
If we want to pick the most likely hypothesis H*, we can drop P(E)
Posterior probability of Hi Prior probability of Hi
Likelihood of data/evidenceif Hi is true
)()|()|( iii HPHEPEHP
2008 © ChengXiang Zhai Dragon Star Lecture at Beijing University, June 21-30, 2008 24
Random Variable• X: S (“measure” of outcome)
– E.g., number of heads, all same face?, …
• Events can be defined according to X
– E(X=a) = {si|X(si)=a}
– E(Xa) = {si|X(si) a}
• So, probabilities can be defined on X
– P(X=a) = P(E(X=a))
– P(aX) = P(E(aX))
• Discrete vs. continuous random variable (think of “partitioning the sample space”)
2008 © ChengXiang Zhai Dragon Star Lecture at Beijing University, June 21-30, 2008 25
An Example: Doc Classification
X1: [sport 1 0 1 1]
Topic the computer game baseball
X2: [sport 1 1 1 1]
X3: [computer 1 1 0 0]
X4: [computer 1 1 1 0]
X5: [other 0 0 1 1] … …
For 3 topics, four words, n=?
Events
Esport ={xi | topic(xi )=“sport”}
Ebaseball ={xi | baseball(xi )=1}
Ebaseball,computer = {xi | baseball(xi )=1 & computer(xi )=0}
Sample Space S={x1,…, xn}
Conditional Probabilities:P(Esport | Ebaseball ), P(Ebaseball|Esport),
P(Esport | Ebaseball, computer ), ...
An inference problem:
Suppose we observe that “baseball” is mentioned, how likely the topic is about “sport”?
But, P(B=1|T=“sport”)=?, P(T=“sport” )=?
P(T=“sport”|B=1) P(B=1|T=“sport”)P(T=“sport”)
Thinking in terms of random variables
Topic: T {“sport”, “computer”, “other”}, “Baseball”: B {0,1}, … P(T=“sport”|B=1), P(B=1|T=“sport”), ...
2008 © ChengXiang Zhai Dragon Star Lecture at Beijing University, June 21-30, 2008 26
Getting to Statistics ...
• P(B=1|T=“sport”)=? (parameter estimation)
– If we see the results of a huge number of random experiments, then
– But, what if we only see a small sample (e.g., 2)? Is this estimate still reliable?
• In general, statistics has to do with drawing conclusions on the whole population based on observations of a sample (data)
)""(
)"",1()""|1(ˆ
sportTcount
sportTBcountsportTBP
2008 © ChengXiang Zhai Dragon Star Lecture at Beijing University, June 21-30, 2008 27
Parameter Estimation
• General setting:
– Given a (hypothesized & probabilistic) model that governs the random experiment
– The model gives a probability of any data p(D|) that depends on the parameter
– Now, given actual sample data X={x1,…,xn}, what can we say about the value of ?
• Intuitively, take your best guess of -- “best” means “best explaining/fitting the data”
• Generally an optimization problem
2008 © ChengXiang Zhai Dragon Star Lecture at Beijing University, June 21-30, 2008 28
Maximum Likelihood vs. Bayesian
• Maximum likelihood estimation
– “Best” means “data likelihood reaches maximum”
– Problem: small sample
• Bayesian estimation
– “Best” means being consistent with our “prior” knowledge and explaining data well
– Problem: how to define prior?
)|(maxargˆ
XP
)()|(maxarg)|(maxargˆ
PXPXP
2008 © ChengXiang Zhai Dragon Star Lecture at Beijing University, June 21-30, 2008 29
Illustration of Bayesian Estimation
Prior: p()
Likelihood: p(X|)
X=(x1,…,xN)
Posterior: p(|X) p(X|)p()
: prior mode ml: ML estimate: posterior mode
2008 © ChengXiang Zhai Dragon Star Lecture at Beijing University, June 21-30, 2008 30
Maximum Likelihood Estimate
Data: a document d with counts c(w1), …, c(wN), and length |d|Model: multinomial distribution M with parameters {p(wi)} Likelihood: p(d|M)Maximum likelihood estimator: M=argmax M p(d|M)
( ) ( )
1 11
1
'
1 1
'
1 1
| |( | ) , ( )
( )... ( )
( | ) log ( | ) ( ) log
( | ) ( ) log ( 1)
( ) ( )0
1, ( ) | |
i i
N Nc w c w
i i i ii iN
N
i ii
N N
i i ii i
i ii
i i
N N
i ii i
dp d M where p w
c w c w
l d M p d M c w
l d M c w
c w c wl
Since c w d So
( ), ( )
| |i
i i
c wp w
d
We’ll tune p(wi) to maximize l(d|M)
Use Lagrange multiplier approach
Set partial derivatives to zero
ML estimate
11
N
ii
2008 © ChengXiang Zhai Dragon Star Lecture at Beijing University, June 21-30, 2008 31
What You Should Know
• Probability concepts:
– sample space, event, random variable, conditional prob. multinomial distribution, etc
• Bayes formula and its interpretation
• Statistics: Know how to compute maximum likelihood estimate
2008 © ChengXiang Zhai Dragon Star Lecture at Beijing University, June 21-30, 2008 32
Essential Background 2:
Basic Concepts in Information Theory
2008 © ChengXiang Zhai Dragon Star Lecture at Beijing University, June 21-30, 2008 33
Information Theory
• Developed by Shannon in the 40s
• Maximizing the amount of information that can be transmitted over an imperfect communication channel
• Data compression (entropy)
• Transmission rate (channel capacity)
2008 © ChengXiang Zhai Dragon Star Lecture at Beijing University, June 21-30, 2008 34
Basic Concepts in Information Theory
• Entropy: Measuring uncertainty of a random variable
• Kullback-Leibler divergence: comparing two distributions
• Mutual Information: measuring the correlation of two random variables
2008 © ChengXiang Zhai Dragon Star Lecture at Beijing University, June 21-30, 2008 35
Entropy: Motivation
• Feature selection: – If we use only a few words to classify docs, what kind of
words should we use?
– P(Topic| “computer”=1) vs p(Topic | “the”=1): which is more random?
• Text compression: – Some documents (less random) can be compressed more than
others (more random)
– Can we quantify the “compressibility”?
• In general, given a random variable X following distribution p(X), – How do we measure the “randomness” of X?
– How do we design optimal coding for X?
2008 © ChengXiang Zhai Dragon Star Lecture at Beijing University, June 21-30, 2008 36
Entropy: Definition
2
( ) ( ) ( ) log ( )
0 log 0 0, log logx
H X H p p x p x all possible values
Define
Entropy H(X) measures the uncertainty/randomness of random variable X
1 ( ) 0.5
( ) 0 1 ( ) 0.8
0 ( ) 1
fair coin p Head
H X between and biased coin p Head
completely biased p Head
Example:
P(Head)
H(X)
1.0
2008 © ChengXiang Zhai Dragon Star Lecture at Beijing University, June 21-30, 2008 37
Entropy: Properties
• Minimum value of H(X): 0
– What kind of X has the minimum entropy?
• Maximum value of H(X): log M, where M is the number of possible values for X
– What kind of X has the maximum entropy?
• Related to coding
p(x)
1log E
p(x)
1p(x)log
p(x)p(x)logH(X)
2
x2
x2
" " "# " log ( ) ( ) [ log ( )]pInformation of x bits to code x p x H X E p x
2008 © ChengXiang Zhai Dragon Star Lecture at Beijing University, June 21-30, 2008 38
Interpretations of H(X)
• Measures the “amount of information” in X
– Think of each value of X as a “message”
– Think of X as a random experiment (20 questions)
• Minimum average number of bits to compress values of X
– The more random X is, the harder to compress
A fair coin has the maximum information, and is hardest to compressA biased coin has some information, and can be compressed to <1 bit on averageA completely biased coin has no information, and needs only 0 bit
" " "# " log ( ) ( ) [ log ( )]pInformation of x bits to code x p x H X E p x
2008 © ChengXiang Zhai Dragon Star Lecture at Beijing University, June 21-30, 2008 39
Conditional Entropy
• The conditional entropy of a random variable Y given another X, expresses how much extra information one still needs to supply on average to communicate Y given that the other party knows X
• H(Topic| “computer”) vs. H(Topic | “the”)?
X)|p(Y logE x)|p(y log y)p(x,
x)|p(y log x)|p(yp(x)
x)X|p(x)H(YX)|H(Y
X Y
X Y
X
x y
x y
x
2008 © ChengXiang Zhai Dragon Star Lecture at Beijing University, June 21-30, 2008 40
Cross Entropy H(p,q)
What if we encode X with a code optimized for a wrong distribution q?
Expected # of bits=? ( , ) [ log ( )] ( ) log ( )px
H p q E q x p x q x
Intuitively, H(p,q) H(p), and mathematically,
( )( , ) ( ) ( )[ log ]
( )
( )log [ ( ) ] 0
( )
x
x
q xH p q H p p x
p x
q xp x
p x
' : ( ) ( )
, , 1
i i i ii i
ii
By Jensen s inequality p f x f p x
where f is a convex function and p
2008 © ChengXiang Zhai Dragon Star Lecture at Beijing University, June 21-30, 2008 41
Kullback-Leibler Divergence D(p||q)
What if we encode X with a code optimized for a wrong distribution q?
How many bits would we waste? ( )
( || ) ( , ) ( ) ( ) log( )x
p xD p q H p q H p p x
q x
Properties:
- D(p||q)0 - D(p||q)D(q||p) - D(p||q)=0 iff p=q
KL-divergence is often used to measure the distance between two distributions
Interpretation:
-Fix p, D(p||q) and H(p,q) vary in the same way
-If p is an empirical distribution, minimize D(p||q) or H(p,q) is equivalent to maximizing likelihood
Relative entropy
2008 © ChengXiang Zhai Dragon Star Lecture at Beijing University, June 21-30, 2008 42
Cross Entropy, KL-Div, and Likelihood
1
1
/ : ( ,..., )
11: ( ) ( , ) ( , )
0 . .
N
N
ii
Data Sample for X Y y y
if x yEmpirical distribution p x y x y x
o wN
1
1
( ) ( )
log ( ) log ( ) ( ) log ( ) ( ) log ( )
N
ii
N
ii x x
L Y p X y
L Y p X y c x p X x N p x p x
Likelihood:
log Likelihood:
1log ( )
1log ( ) ( , ) ( || ) ( )
1, arg max log ( ) arg min ( , ) arg min ( , ) arg min 2
L YN
p p p p
L Y H p p D p p H pN
Fix the data L Y H p p D p pN
Criterion for selecting a good model Perplexity(p)
Nxcxp /)()(~
2008 © ChengXiang Zhai Dragon Star Lecture at Beijing University, June 21-30, 2008 43
Mutual Information I(X;Y)
Comparing two distributions: p(x,y) vs p(x)p(y)
,
( , )( ; ) ( , ) log ( ) ( | ) ( ) ( | )
( ) ( )x y
p x yI X Y p x y H X H X Y H Y H Y X
p x p y
Properties: I(X;Y)0; I(X;Y)=I(Y;X); I(X;Y)=0 iff X & Y are independent
Interpretations: - Measures how much reduction in uncertainty of X given info. about Y - Measures correlation between X and Y - Related to the “channel capacity” in information theory
Examples: I(Topic; “computer”) vs. I(Topic; “the”)?
I(“computer”, “program”) vs (“computer”, “baseball”)?
2008 © ChengXiang Zhai Dragon Star Lecture at Beijing University, June 21-30, 2008 44
What You Should Know
• Information theory concepts: entropy, cross entropy, relative entropy, conditional entropy, KL-div., mutual information
– Know their definitions, how to compute them
– Know how to interpret them
– Know their relationships
2008 © ChengXiang Zhai Dragon Star Lecture at Beijing University, June 21-30, 2008 45
Essential Background 3:
Natural Language Processing
2008 © ChengXiang Zhai Dragon Star Lecture at Beijing University, June 21-30, 2008 46
What is NLP?
)ِه) … -ْه+ِل َأ َو-َم-َع- ِه) -ْف+ِس) َن َم-َع- 5 َو-َص-ًاِد)َق-ًا 5 -ًا +َن َأَم)ْي <وَن- -ُك َي َأَن ًاَن) +ِس- اإلَن َع-ِل-ى -ِج)ُب< َيَع-ِل-ى -ْع+َم-َل- َي -َن+ َو-َأ الو-َط-ِن) َن)
+ ْأ َش- )َع+الِء) ِإ ِف)ي Tُج<ْه+ٍد Xَل> ُك +ُذ<َل- -ْب َي -َن+ َو-َأ )ِه) اَن +َر- ْي َو-ُج)… َم-ًا
How can a computer make sense out of such a string?
- What are the basic units of meaning (words)?- What is the meaning of each word? - How are words related with each other? - What is the “combined meaning” of words? - What is the “meta-meaning”? (speech act)- Handling a large chunk of text- Making sense of everything
Syntax
Semantics
Pragmatics
Morphology
DiscourseInference
La listas actualizadas figuran como Aneio I.
Arabic text
Spanish text
2008 © ChengXiang Zhai Dragon Star Lecture at Beijing University, June 21-30, 2008 47
An Example of NLP
A dog is chasing a boy on the playgroundDet Noun Aux Verb Det Noun Prep Det Noun
Noun Phrase Complex Verb Noun PhraseNoun Phrase
Prep PhraseVerb Phrase
Verb Phrase
Sentence
Dog(d1).Boy(b1).Playground(p1).Chasing(d1,b1,p1).
Semantic analysis
Lexicalanalysis
(part-of-speechtagging)
Syntactic analysis(Parsing)
A person saying this maybe reminding another person to
get the dog back…
Pragmatic analysis(speech act)
Scared(x) if Chasing(_,x,_).+
Scared(b1)
Inference
2008 © ChengXiang Zhai Dragon Star Lecture at Beijing University, June 21-30, 2008 48
If we can do this for all the sentences, then …
BAD NEWS:
Unfortunately, we can’t.
General NLP = “AI-Complete”
2008 © ChengXiang Zhai Dragon Star Lecture at Beijing University, June 21-30, 2008 49
NLP is Difficult!
• Natural language is designed to make human communication efficient. As a result,
– we omit a lot of “common sense” knowledge, which we assume the hearer/reader possesses
– we keep a lot of ambiguities, which we assume the hearer/reader knows how to resolve
• This makes EVERY step in NLP hard
– Ambiguity is a “killer”
– Common sense reasoning is pre-required
2008 © ChengXiang Zhai Dragon Star Lecture at Beijing University, June 21-30, 2008 50
Examples of Challenges• Word-level ambiguity: E.g.,
– “design” can be a noun or a verb (Ambiguous POS)
– “root” has multiple meanings (Ambiguous sense)
• Syntactic ambiguity: E.g.,
– “natural language processing” (Modification)
– “A man saw a boy with a telescope.” (PP Attachment)
• Anaphora resolution: “John persuaded Bill to buy a TV for himself.” (himself = John or Bill?)
• Presupposition: “He has quit smoking.” implies that he smoked before.
2008 © ChengXiang Zhai Dragon Star Lecture at Beijing University, June 21-30, 2008 51
Despite all the challenges, research in NLP has also made
a lot of progress…
2008 © ChengXiang Zhai Dragon Star Lecture at Beijing University, June 21-30, 2008 52
High-level History of NLP• Early enthusiasm (1950’s): Machine Translation
– Too ambitious
– Bar-Hillel report (1960) concluded that fully-automatic high-quality translation could not be accomplished without knowledge (Dictionary + Encyclopedia)
• Less ambitious applications (late 1960’s & early 1970’s): Limited success, failed to scale up
– Speech recognition
– Dialogue (Eliza)
– Inference and domain knowledge (SHRDLU=“block world”)
• Real world evaluation (late 1970’s – now)– Story understanding (late 1970’s & early 1980’s)
– Large scale evaluation of speech recognition, text retrieval, information extraction (1980 – now)
– Statistical approaches enjoy more success (first in speech recognition & retrieval, later others)
• Current trend: – Boundary between statistical and symbolic approaches is disappearing.
– We need to use all the available knowledge
– Application-driven NLP research (bioinformatics, Web, Question answering…)
Stat. language models
Robust component techniques
Applications
Knowledge representation
Deep understanding in
limited domainShallow understanding
2008 © ChengXiang Zhai Dragon Star Lecture at Beijing University, June 21-30, 2008 53
The State of the Art
A dog is chasing a boy on the playgroundDet Noun Aux Verb Det Noun Prep Det Noun
Noun Phrase Complex Verb Noun PhraseNoun Phrase
Prep PhraseVerb Phrase
Verb Phrase
Sentence
Semantics: some aspects
- Entity/relation extraction- Word sense disambiguation- Anaphora resolution
POSTagging:
97%
Parsing: partial >90%(?)
Speech act analysis: ???
Inference: ???
2008 © ChengXiang Zhai Dragon Star Lecture at Beijing University, June 21-30, 2008 54
Technique Showcase: POS Tagging
This sentence serves as an example of Det N V1 P Det N P annotated text… V2 N
Training data (Annotated text)
POS Tagger“This is a new sentence”This is a new sentenceDet Aux Det Adj N
This is a new sentenceDet Det Det Det Det
… …Det Aux Det Adj N … …V2 V2 V2 V2 V2
Consider all possibilities,and pick the one withthe highest probability
1 1
1 1 1
11
( ,..., , ,..., )
( | )... ( | ) ( )... ( )
( | ) ( | )
k k
k k k
k
i i i ii
p w w t t
p t w p t w p w p w
p w t p t t
Method 1: Independent assignmentMost common tag
Method 2: Partial dependency
2008 © ChengXiang Zhai Dragon Star Lecture at Beijing University, June 21-30, 2008 55
Technique Showcase: ParsingS NP VPNP Det BNPNP BNPNP NP PPBNP NVP V VP Aux V NPVP VP PPPP P NP
V chasingAux isN dogN boyN playgroundDet theDet aP on
Grammar
Lexicon
Generate
S
NP VP
BNP
N
Det
A
dog
VP PP
Aux V
is
the playground
ona boy
chasing
NP P NP
S
NP VP
BNP
N
dog
PPAux V
is
ona boy
chasing
NP
P NP
Det
A
the playground
NP
roller skates
1.00.30.40.3
1.0
…
…
0.01
0.003
…
…
Probability of this tree=0.000015
Choose a tree with highest prob….
Can also be treated as a classification/decision problem…
2008 © ChengXiang Zhai Dragon Star Lecture at Beijing University, June 21-30, 2008 56
Semantic Analysis Techniques
• Only successful for VERY limited domain or for SOME aspect of semantics
• E.g.,
– Entity extraction (e.g., recognizing a person’s name): Use rules and/or machine learning
– Word sense disambiguation: addressed as a classification problem with supervised learning
– Anaphora resolution …
In general, exploiting machine learning
and statistical language models…
2008 © ChengXiang Zhai Dragon Star Lecture at Beijing University, June 21-30, 2008 57
What We Can’t Do
• 100% POS tagging– “He turned off the highway.” vs “He turned off the fan.”
• General complete parsing– “A man saw a boy with a telescope.”
• Deep semantic analysis– Will we ever be able to precisely define the meaning of “own” in “John
owns a restaurant.”?
Robust & general NLP tends to be “shallow”,
while “deep” understanding doesn’t scale up …
2008 © ChengXiang Zhai Dragon Star Lecture at Beijing University, June 21-30, 2008 58
Major NLP Applications
• Speech recognition: e.g., Auto telephone call routing
• Text management– Text retrieval/filtering
– Text classification
– Text summarization
– Text mining
– Query answering
• Language tutoring– Spelling/grammar correction
• Machine translation– Cross-language retrieval
– Restricted natural language
• Natural language user interface
Our focus
2008 © ChengXiang Zhai Dragon Star Lecture at Beijing University, June 21-30, 2008 59
NLP & Text Management
Better NLP => Better Text Management
Bad NLP => Bad Text Management?
Robust, shallow NLP tends to be more useful than deep, but fragile NLP.
Errors in NLP can hurt text management performance…
2008 © ChengXiang Zhai Dragon Star Lecture at Beijing University, June 21-30, 2008 60
How Much NLP is Really Needed?
Tasks Dependency on NLP
Classification/Retrieval
Summarization/
Extraction/
Mining
Translation/Dialogue
Question Answering
“Easier” & More “workarounds”
2008 © ChengXiang Zhai Dragon Star Lecture at Beijing University, June 21-30, 2008 61
Workaround Example I: ELIZA
Person : Men are all alike.ELIZA : In what way?Person : They're always bugging us about something or other.ELIZA : Can you think of a specific example?Person : Well, my boyfriend made me come here.ELIZA : Your boyfriend made you come here?Person : He says I'm depressed much of the time.ELIZA : I am sorry to hear that you are depressed.Person : It's true. I am unhappy.ELIZA : Do you think coming here will help you not to be unhappy?Person : I need some help, that much seems certain.ELIZA : What would it mean to you if you got some help?Person : Perhaps I could learn to get along with my mother.ELIZA : Tell me more about your family
ELIZA plays the role of a therapistHow does it work?
Heuristic rules based on Keyword matching! E.g.,
I remember X =>Do you often think of X?
always => Can you think of a specific example?
It keeps the dialogue going on, but NEVER makes any claim!
A breakdown example:Person : Failure is the mother of success.ELIZA : Tell me more about your family
No NLP, but useful. Perhaps we should call this NLP?Statistical NLP often has a similar flavor with “SOFT” rules LEARNED from data
2008 © ChengXiang Zhai Dragon Star Lecture at Beijing University, June 21-30, 2008 62
Workaround Example II: Statistical Translation
Learn how to translate Chinese to English from many example translations
Intuitions:
- If we have seen all possible translations, then we simply lookup- If we have seen a similar translation, then we can adapt- If we haven’t seen any example that’s similar, we try to generalize what we’ve seen
EnglishSpeaker
TranslatorNoisyChannel
P(E) P(C|E)
EnglishWords (E)
ChineseWords(C)
EnglishTranslation
P(E|C)=?
All these intuitions are captured through a probabilistic model
2008 © ChengXiang Zhai Dragon Star Lecture at Beijing University, June 21-30, 2008 63
So, what NLP techniques are most useful for text management?
Statistical NLP in general, and
statistical language models in particular
The need for high robustness and efficiency implies the dominant use of
simple models (i.e., unigram models)
2008 © ChengXiang Zhai Dragon Star Lecture at Beijing University, June 21-30, 2008 64
What You Should Know• NLP is the basis for text management
– Better NLP enables better text management
– Better NLP is necessary for sophisticated tasks
• But – Bad NLP doesn’t mean bad text management
– There are often “workarounds” for a task
– Inaccurate NLP can even hurt the performance of a task
• The most effective NLP techniques are often statistical with the help of linguistic knowledge
• The challenge is to bridge the gap between NLP and applications
2008 © ChengXiang Zhai Dragon Star Lecture at Beijing University, June 21-30, 2008 65
Roadmap
• Today’s lecture
– Course overview
– Essential background (prob & stat, info theory, NLP)
• Next two lectures: overview of IR
– Basic concepts
– Evaluation
– Brief history
– Basic models
– …
top related