SlideShare a Scribd company logo

Natural Language Processing:Part-Of-Speech Tagging, Sequence Labeling, and Hidden Markov Models (HMMs).ppt

Download as PPT, PDF
S
saimakhosa3

Natural Language Processing:Part-Of-Speech Tagging,Sequence Labeling, andHidden Markov Models (HMMs)

Technology
1 of 84
Download to read offline
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
1 1 CS 388: Natural Language Processing: Part-Of-Speech Tagging, Sequence Labeling, and Hidden Markov Models (HMMs) Raymond J. Mooney University of Texas at Austin
2 Part Of Speech Tagging • Annotate each word in a sentence with a part-of-speech marker. • Lowest level of syntactic analysis. • Useful for subsequent syntactic parsing and word sense disambiguation. John saw the saw and decided to take it to the table. NNP VBD DT NN CC VBD TO VB PRP IN DT NN
3 English POS Tagsets • Original Brown corpus used a large set of 87 POS tags. • Most common in NLP today is the Penn Treebank set of 45 tags. – Tagset used in these slides. – Reduced from the Brown set for use in the context of a parsed corpus (i.e. treebank). • The C5 tagset used for the British National Corpus (BNC) has 61 tags.
4 English Parts of Speech • Noun (person, place or thing) – Singular (NN): dog, fork – Plural (NNS): dogs, forks – Proper (NNP, NNPS): John, Springfields – Personal pronoun (PRP): I, you, he, she, it – Wh-pronoun (WP): who, what • Verb (actions and processes) – Base, infinitive (VB): eat – Past tense (VBD): ate – Gerund (VBG): eating – Past participle (VBN): eaten – Non 3rd person singular present tense (VBP): eat – 3rd person singular present tense: (VBZ): eats – Modal (MD): should, can – To (TO): to (to eat)
5 English Parts of Speech (cont.) • Adjective (modify nouns) – Basic (JJ): red, tall – Comparative (JJR): redder, taller – Superlative (JJS): reddest, tallest • Adverb (modify verbs) – Basic (RB): quickly – Comparative (RBR): quicker – Superlative (RBS): quickest • Preposition (IN): on, in, by, to, with • Determiner: – Basic (DT) a, an, the – WH-determiner (WDT): which, that • Coordinating Conjunction (CC): and, but, or, • Particle (RP): off (took off), up (put up)
Closed vs. Open Class • Closed class categories are composed of a small, fixed set of grammatical function words for a given language. – Pronouns, Prepositions, Modals, Determiners, Particles, Conjunctions • Open class categories have large number of words and new ones are easily invented. – Nouns (Googler, textlish), Verbs (Google), Adjectives (geeky), Abverb (automagically) 6
7 Ambiguity in POS Tagging • “Like” can be a verb or a preposition – I like/VBP candy. – Time flies like/IN an arrow. • “Around” can be a preposition, particle, or adverb – I bought it at the shop around/IN the corner. – I never got around/RP to getting a car. – A new Prius costs around/RB $25K.
8 POS Tagging Process • Usually assume a separate initial tokenization process that separates and/or disambiguates punctuation, including detecting sentence boundaries. • Degree of ambiguity in English (based on Brown corpus) – 11.5% of word types are ambiguous. – 40% of word tokens are ambiguous. • Average POS tagging disagreement amongst expert human judges for the Penn treebank was 3.5% – Based on correcting the output of an initial automated tagger, which was deemed to be more accurate than tagging from scratch. • Baseline: Picking the most frequent tag for each specific word type gives about 90% accuracy – 93.7% if use model for unknown words for Penn Treebank tagset.
9 POS Tagging Approaches • Rule-Based: Human crafted rules based on lexical and other linguistic knowledge. • Learning-Based: Trained on human annotated corpora like the Penn Treebank. – Statistical models: Hidden Markov Model (HMM), Maximum Entropy Markov Model (MEMM), Conditional Random Field (CRF) – Rule learning: Transformation Based Learning (TBL) – Neural networks: Recurrent networks like Long Short Term Memory (LSTMs) • Generally, learning-based approaches have been found to be more effective overall, taking into account the total amount of human expertise and effort involved.
10 Classification Learning • Typical machine learning addresses the problem of classifying a feature-vector description into a fixed number of classes. • There are many standard learning methods for this task: – Decision Trees and Rule Learning – Naïve Bayes and Bayesian Networks – Logistic Regression / Maximum Entropy (MaxEnt) – Perceptron and Neural Networks – Support Vector Machines (SVMs) – Nearest-Neighbor / Instance-Based
11 Beyond Classification Learning • Standard classification problem assumes individual cases are disconnected and independent (i.i.d.: independently and identically distributed). • Many NLP problems do not satisfy this assumption and involve making many connected decisions, each resolving a different ambiguity, but which are mutually dependent. • More sophisticated learning and inference techniques are needed to handle such situations in general.
12 Sequence Labeling Problem • Many NLP problems can viewed as sequence labeling. • Each token in a sequence is assigned a label. • Labels of tokens are dependent on the labels of other tokens in the sequence, particularly their neighbors (not i.i.d). foo bar blam zonk zonk bar blam
13 Information Extraction • Identify phrases in language that refer to specific types of entities and relations in text. • Named entity recognition is task of identifying names of people, places, organizations, etc. in text. people organizations places – Michael Dell is the CEO of Dell Computer Corporation and lives in Austin Texas. • Extract pieces of information relevant to a specific application, e.g. used car ads: make model year mileage price – For sale, 2002 Toyota Prius, 20,000 mi, $15K or best offer. Available starting July 30, 2006.
14 Semantic Role Labeling • For each clause, determine the semantic role played by each noun phrase that is an argument to the verb. agent patient source destination instrument – John drove Mary from Austin to Dallas in his Toyota Prius. – The hammer broke the window. • Also referred to a “case role analysis,” “thematic analysis,” and “shallow semantic parsing”
15 Bioinformatics • Sequence labeling also valuable in labeling genetic sequences in genome analysis. extron intron – AGCTAACGTTCGATACGGATTACAGCCT
16 Problems with Sequence Labeling as Classification • Not easy to integrate information from category of tokens on both sides. • Difficult to propagate uncertainty between decisions and “collectively” determine the most likely joint assignment of categories to all of the tokens in a sequence.
17 Probabilistic Sequence Models • Probabilistic sequence models allow integrating uncertainty over multiple, interdependent classifications and collectively determine the most likely global assignment. • Two standard models – Hidden Markov Model (HMM) – Conditional Random Field (CRF)
18 Markov Model / Markov Chain • A finite state machine with probabilistic state transitions. • Makes Markov assumption that next state only depends on the current state and independent of previous history.
19 Sample Markov Model for POS 0.95 0.9 0.05 stop 0.5 0.1 0.8 0.1 0.1 0.25 0.25 start 0.1 0.5 0.4 Det Noun PropNoun Verb
20 Sample Markov Model for POS 0.95 0.9 0.05 stop 0.5 0.1 0.8 0.1 0.1 0.25 0.25 start 0.1 0.5 0.4 Det Noun PropNoun Verb P(PropNoun Verb Det Noun) = 0.4*0.8*0.25*0.95*0.1=0.0076
21 Hidden Markov Model • Probabilistic generative model for sequences. • Assume an underlying set of hidden (unobserved, latent) states in which the model can be (e.g. parts of speech). • Assume probabilistic transitions between states over time (e.g. transition from POS to another POS as sequence is generated). • Assume a probabilistic generation of tokens from states (e.g. words generated for each POS).
22 Sample HMM for POS PropNoun JohnMary Alice Jerry Tom Noun cat dog car pen bed apple Det a the th e the that a the a Verb bit ate saw played hit 0.95 0.9 gave 0.05 stop 0.5 0.1 0.8 0.1 0.1 0.25 0.25 start 0.1 0.5 0.4
23 Sample HMM Generation PropNoun JohnMary Alice Jerry Tom Noun cat dog car pen bed apple Det a the th e the that a the a Verb bit ate saw played hit 0.95 0.9 gave 0.05 stop 0.5 0.1 0.8 0.1 0.1 0.25 0.25 start 0.1 0.5 0.4
24 Sample HMM Generation PropNoun JohnMary Alice Jerry Tom Noun cat dog car pen bed apple Det a the th e the that a the a Verb bit ate saw played hit 0.95 0.9 gave 0.05 stop 0.5 0.1 0.8 0.1 0.1 0.25 start 0.1 0.5 0.4
25 Sample HMM Generation PropNoun JohnMary Alice Jerry Tom Noun cat dog car pen bed apple Det a the th e the that a the a Verb bit ate saw played hit 0.95 0.9 gave 0.05 stop 0.5 0.1 0.8 0.1 0.1 0.25 0.25 John start 0.1 0.5 0.4
26 Sample HMM Generation PropNoun JohnMary Alice Jerry Tom Noun cat dog car pen bed apple Det a the th e the that a the a Verb bit ate saw played hit 0.95 0.9 gave 0.05 stop 0.5 0.1 0.8 0.1 0.1 0.25 0.25 John start 0.1 0.5 0.4
27 Sample HMM Generation PropNoun JohnMary Alice Jerry Tom Noun cat dog car pen bed apple Det a the th e the that a the a Verb bit ate saw played hit 0.95 0.9 gave 0.05 stop 0.5 0.1 0.8 0.1 0.1 0.25 0.25 John bit start 0.1 0.5 0.4
28 Sample HMM Generation PropNoun JohnMary Alice Jerry Tom Noun cat dog car pen bed apple Det a the th e the that a the a Verb bit ate saw played hit 0.95 0.9 gave 0.05 stop 0.5 0.1 0.8 0.1 0.1 0.25 0.25 John bit start 0.1 0.5 0.4
29 Sample HMM Generation PropNoun JohnMary Alice Jerry Tom Noun cat dog car pen bed apple Det a the th e the that a the a Verb bit ate saw played hit 0.95 0.9 gave 0.05 stop 0.5 0.1 0.8 0.1 0.1 0.25 0.25 John bit the start 0.1 0.5 0.4
30 Sample HMM Generation PropNoun JohnMary Alice Jerry Tom Noun cat dog car pen bed apple Det a the th e the that a the a Verb bit ate saw played hit 0.95 0.9 gave 0.05 stop 0.5 0.1 0.8 0.1 0.1 0.25 0.25 John bit the start 0.1 0.5 0.4
31 Sample HMM Generation PropNoun JohnMary Alice Jerry Tom Noun cat dog car pen bed apple Det a the th e the that a the a Verb bit ate saw played hit 0.95 0.9 gave 0.05 stop 0.5 0.1 0.8 0.1 0.1 0.25 0.25 John bit the apple start 0.1 0.5 0.4
32 Sample HMM Generation PropNoun JohnMary Alice Jerry Tom Noun cat dog car pen bed apple Det a the th e the that a the a Verb bit ate saw played hit 0.95 0.9 gave 0.05 stop 0.5 0.1 0.8 0.1 0.1 0.25 0.25 John bit the apple start 0.1 0.5 0.4
33 Formal Definition of an HMM • A set of N +2 states S={s0,s1,s2, … sN, sF} – Distinguished start state: s0 – Distinguished final state: sF • A set of M possible observations V={v1,v2…vM} • A state transition probability distribution A={aij} • Observation probability distribution for each state j B={bj(k)} • Total parameter set λ={A,B} F j i N j i s q s q P a i t j t ij         , 0 and , 1 ) | ( 1 M k 1 1 ) | at ( ) (       N j s q t v P k b j t k j N i a a iF N j ij       0 1 1
34 HMM Generation Procedure • To generate a sequence of T observations: O = o1 o2 … oT Set initial state q1=s0 For t = 1 to T Transit to another state qt+1=sj based on transition distribution aij for state qt Pick an observation ot=vk based on being in state qt using distribution bqt(k)
35 Three Useful HMM Tasks • Observation Likelihood: To classify and order sequences. • Most likely state sequence (Decoding): To tag each token in a sequence with a label. • Maximum likelihood training (Learning): To train models to fit empirical training data.
36 HMM: Observation Likelihood • Given a sequence of observations, O, and a model with a set of parameters, λ, what is the probability that this observation was generated by this model: P(O| λ) ? • Allows HMM to be used as a language model: A formal probabilistic model of a language that assigns a probability to each string saying how likely that string was to have been generated by the language. • Useful for two tasks: – Sequence Classification – Most Likely Sequence
37 Sequence Classification • Assume an HMM is available for each category (i.e. language). • What is the most likely category for a given observation sequence, i.e. which category’s HMM is most likely to have generated it? • Used in speech recognition to find most likely word model to have generate a given sound or phoneme sequence. Austin Boston ? ? P(O | Austin) > P(O | Boston) ? ah s t e n O
38 Most Likely Sequence • Of two or more possible sequences, which one was most likely generated by a given model? • Used to score alternative word sequence interpretations in speech recognition. Ordinary English dice precedent core vice president Gore O1 O2 ? ? P(O2 | OrdEnglish) > P(O1 | OrdEnglish) ?
39 HMM: Observation Likelihood Naïve Solution • Consider all possible state sequences, Q, of length T that the model could have traversed in generating the given observation sequence. • Compute the probability of a given state sequence from A, and multiply it by the probabilities of generating each of given observations in each of the corresponding states in this sequence to get P(O,Q| λ) = P(O| Q, λ) P(Q| λ) . • Sum this over all possible state sequences to get P(O| λ). • Computationally complex: O(TNT ).
40 HMM: Observation Likelihood Efficient Solution • Due to the Markov assumption, the probability of being in any state at any given time t only relies on the probability of being in each of the possible states at time t−1. • Forward Algorithm: Uses dynamic programming to exploit this fact to efficiently compute observation likelihood in O(TN2 ) time. – Compute a forward trellis that compactly and implicitly encodes information about all possible state paths.
Forward Trellis 41 s1 s2 sN       s0 sF                      t1 t2 t3 tT-1 tT • Continue forward in time until reaching final time point and sum probability of ending in final state.
Forward Probabilities • Let t(j) be the probability of being in state j after seeing the first t observations (by summing over all initial paths leading to j). 42 ) | , ,... , ( ) ( 2 1   j t t t s q o o o P j  
Forward Step 43 s1 s2 sN    sj t-1(i) t(i) a1j a2j aNj a2j • Consider all possible ways of getting to sj at time t by coming from all possible states si and determine probability of each. • Sum these to get the total probability of being in state sj at time t while accounting for the first t −1 observations. • Then multiply by the probability of actually observing ot in sj.
Computing the Forward Probabilities • Initialization • Recursion • Termination 44 N j o b a j j j    1 ) ( ) ( 1 0 1  T t N j o b a i j t j N i ij t t               1 , 1 ) ( ) ( ) ( 1 1        N i iF T F T a i s O P 1 1 ) ( ) ( ) | (   
Forward Computational Complexity • Requires only O(TN2 ) time to compute the probability of an observed sequence given a model. • Exploits the fact that all state sequences must merge into one of the N possible states at any point in time and the Markov assumption that only the last state effects the next one. 45
46 Most Likely State Sequence (Decoding) • Given an observation sequence, O, and a model, λ, what is the most likely state sequence,Q=q1,q2,…qT, that generated this sequence from this model? • Used for sequence labeling, assuming each state corresponds to a tag, it determines the globally best assignment of tags to all tokens in a sequence using a principled approach grounded in probability theory. John gave the dog an apple.
47 Most Likely State Sequence • Given an observation sequence, O, and a model, λ, what is the most likely state sequence,Q=q1,q2,…qT, that generated this sequence from this model? • Used for sequence labeling, assuming each state corresponds to a tag, it determines the globally best assignment of tags to all tokens in a sequence using a principled approach grounded in probability theory. John gave the dog an apple. Det Noun PropNoun Verb
48 Most Likely State Sequence • Given an observation sequence, O, and a model, λ, what is the most likely state sequence,Q=q1,q2,…qT, that generated this sequence from this model? • Used for sequence labeling, assuming each state corresponds to a tag, it determines the globally best assignment of tags to all tokens in a sequence using a principled approach grounded in probability theory. John gave the dog an apple. Det Noun PropNoun Verb
49 Most Likely State Sequence • Given an observation sequence, O, and a model, λ, what is the most likely state sequence,Q=q1,q2,…qT, that generated this sequence from this model? • Used for sequence labeling, assuming each state corresponds to a tag, it determines the globally best assignment of tags to all tokens in a sequence using a principled approach grounded in probability theory. John gave the dog an apple. Det Noun PropNoun Verb
50 Most Likely State Sequence • Given an observation sequence, O, and a model, λ, what is the most likely state sequence,Q=q1,q2,…qT, that generated this sequence from this model? • Used for sequence labeling, assuming each state corresponds to a tag, it determines the globally best assignment of tags to all tokens in a sequence using a principled approach grounded in probability theory. John gave the dog an apple. Det Noun PropNoun Verb
51 Most Likely State Sequence • Given an observation sequence, O, and a model, λ, what is the most likely state sequence,Q=q1,q2,…qT, that generated this sequence from this model? • Used for sequence labeling, assuming each state corresponds to a tag, it determines the globally best assignment of tags to all tokens in a sequence using a principled approach grounded in probability theory. John gave the dog an apple. Det Noun PropNoun Verb
52 Most Likely State Sequence • Given an observation sequence, O, and a model, λ, what is the most likely state sequence,Q=q1,q2,…qT, that generated this sequence from this model? • Used for sequence labeling, assuming each state corresponds to a tag, it determines the globally best assignment of tags to all tokens in a sequence using a principled approach grounded in probability theory. John gave the dog an apple. Det Noun PropNoun Verb
53 HMM: Most Likely State Sequence Efficient Solution • Obviously, could use naïve algorithm based on examining every possible state sequence of length T. • Dynamic Programming can also be used to exploit the Markov assumption and efficiently determine the most likely state sequence for a given observation and model. • Standard procedure is called the Viterbi algorithm (Viterbi, 1967) and also has O(N2 T) time complexity.
Viterbi Scores • Recursively compute the probability of the most likely subsequence of states that accounts for the first t observations and ends in state sj. 54 ) | , ,..., , ,..., , ( max ) ( 1 1 1 0 ,..., , 1 1 0  j t t t q q q t s q o o q q q P j v t     • Also record “backpointers” that subsequently allow backtracing the most probable state sequence.  btt(j) stores the state at time t-1 that maximizes the probability that system was in state sj at time t (given the observed sequence).
Computing the Viterbi Scores • Initialization • Recursion • Termination 55 N j o b a j v j j    1 ) ( ) ( 1 0 1 T t N j o b a i v j v t j ij t N i t        1 , 1 ) ( ) ( max ) ( 1 1 iF T N i F T a i v s v P ) ( max ) ( * 1 1     Analogous to Forward algorithm except take max instead of sum
Computing the Viterbi Backpointers • Initialization • Recursion • Termination 56 N j s j bt    1 ) ( 0 1 T t N j o b a i v j bt t j ij t N i t        1 , 1 ) ( ) ( argmax ) ( 1 1 iF T N i F T T a i v s bt q ) ( argmax ) ( * 1 1     Final state in the most probable state sequence. Follow backpointers to initial state to construct full sequence.
Viterbi Backpointers 57 s1 s2 sN       s0 sF                      t1 t2 t3 tT-1 tT
Viterbi Backtrace 58 s1 s2 sN       s0 sF                      t1 t2 t3 tT-1 tT Most likely Sequence: s0 sN s1 s2 …s2 sF
HMM Learning • Supervised Learning: All training sequences are completely labeled (tagged). • Unsupervised Learning: All training sequences are unlabelled (but generally know the number of tags, i.e. states). • Semisupervised Learning: Some training sequences are labeled, most are unlabeled. 59
60 Supervised HMM Training • If training sequences are labeled (tagged) with the underlying state sequences that generated them, then the parameters, λ={A,B} can all be estimated directly. Supervised HMM Training John ate the apple A dog bit Mary Mary hit the dog John gave Mary the cat. . . . Training Sequences Det Noun PropNoun Verb
Supervised Parameter Estimation • Estimate state transition probabilities based on tag bigram and unigram statistics in the labeled data. • Estimate the observation probabilities based on tag/word co-occurrence statistics in the labeled data. • Use appropriate smoothing if training data is sparse. 61 ) ( ) q , ( 1 t i t j i t ij s q C s s q C a      ) ( ) , ( ) ( j i k i j i j s q C v o s q C k b    
Learning and Using HMM Taggers • Use a corpus of labeled sequence data to easily construct an HMM using supervised training. • Given a novel unlabeled test sequence to tag, use the Viterbi algorithm to predict the most likely (globally optimal) tag sequence. 62
Evaluating Taggers • Train on training set of labeled sequences. • Possibly tune parameters based on performance on a development set. • Measure accuracy on a disjoint test set. • Generally measure tagging accuracy, i.e. the percentage of tokens tagged correctly. • Accuracy of most modern POS taggers, including HMMs is 96−97% (for Penn tagset trained on about 800K words) . – Generally matching human agreement level. 63
64 Unsupervised Maximum Likelihood Training ah s t e n a s t i n oh s t u n eh z t en . . . Training Sequences HMM Training Austin
65 Maximum Likelihood Training • Given an observation sequence, O, what set of parameters, λ, for a given model maximizes the probability that this data was generated from this model (P(O| λ))? • Used to train an HMM model and properly induce its parameters from a set of training data. • Only need to have an unannotated observation sequence (or set of sequences) generated from the model. Does not need to know the correct state sequence(s) for the observation sequence(s). In this sense, it is unsupervised.
Bayes Theorem Simple proof from definition of conditional probability: ) ( ) ( ) | ( ) | ( E P H P H E P E H P  ) ( ) ( ) | ( E P E H P E H P   ) ( ) ( ) | ( H P E H P H E P   ) ( ) | ( ) ( H P H E P E H P   QED: (Def. cond. prob.) (Def. cond. prob.) ) ( ) ( ) | ( ) | ( E P H P H E P E H P 
Maximum Likelihood vs. Maximum A Posteriori (MAP) • The MAP parameter estimate is the most likely given the observed data, O. ) ( ) ( ) | ( argmax ) | ( argmax O P P O P O P MAP         ) ( ) | ( argmax    P O P • If all parameterizations are assumed to be equally likely a priori, then MLE and MAP are the same. • If parameters are given priors (e.g. Gaussian or Lapacian with zero mean), then MAP is a principled way to perform smoothing or regularization.
68 HMM: Maximum Likelihood Training Efficient Solution • There is no known efficient algorithm for finding the parameters, λ, that truly maximizes P(O| λ). • However, using iterative re-estimation, the Baum- Welch algorithm (a.k.a. forward-backward) , a version of a standard statistical procedure called Expectation Maximization (EM), is able to locally maximize P(O| λ). • In practice, EM is able to find a good set of parameters that provide a good fit to the training data in many cases.
69 EM Algorithm • Iterative method for learning probabilistic categorization model from unsupervised data. • Initially assume random assignment of examples to categories. • Learn an initial probabilistic model by estimating model parameters  from this randomly labeled data. • Iterate following two steps until convergence: – Expectation (E-step): Compute P(ci | E) for each example given the current model, and probabilistically re-label the examples based on these posterior probability estimates. – Maximization (M-step): Re-estimate the model parameters, , from the probabilistically re-labeled data.
70 EM Unlabeled Examples +  +  +  +   + Assign random probabilistic labels to unlabeled data Initialize:
71 EM Prob. Learner +  +  +  +   + Give soft-labeled training data to a probabilistic learner Initialize:
72 EM Prob. Learner Prob. Classifier +  +  +  +   + Produce a probabilistic classifier Initialize:
73 EM Prob. Learner Prob. Classifier Relabel unlabled data using the trained classifier +  +  +  +   + E Step:
74 EM Prob. Learner +  +  +  +   + Prob. Classifier Continue EM iterations until probabilistic labels on unlabeled data converge. Retrain classifier on relabeled data M step:
75 Sketch of Baum-Welch (EM) Algorithm for Training HMMs Assume an HMM with N states. Randomly set its parameters λ=(A,B) (making sure they represent legal distributions) Until converge (i.e. λ no longer changes) do: E Step: Use the forward/backward procedure to determine the probability of various possible state sequences for generating the training data M Step: Use these probability estimates to re-estimate values for all of the parameters λ See textbook for detailed equations for E and M steps
EM Properties • Each iteration changes the parameters in a way that is guaranteed to increase the likelihood of the data: P(O|). • Anytime algorithm: Can stop at any time prior to convergence to get approximate solution. • Converges to a local maximum.
Semi-Supervised Learning • EM algorithms can be trained with a mix of labeled and unlabeled data. • EM basically predicts a probabilistic (soft) labeling of the instances and then iteratively retrains using supervised learning on these predicted labels (“self training”). • EM can also exploit supervised data: – 1) Use supervised learning on labeled data to initialize the parameters (instead of initializing them randomly). – 2) Use known labels for supervised data instead of predicting soft labels for these examples during retraining iterations.
78 Semi-Supervised EM Training Examples - - + + + Unlabeled Examples Prob. Learner Prob. Classifier +  +  +  +   +
79 Semi-Supervised EM Training Examples - - + + + Prob. Learner +  +  +  +   + Prob. Classifier
80 Semi-Supervised EM Training Examples - - + + + Prob. Learner +  +  +  +   + Prob. Classifier
81 Semi-Supervised EM Training Examples - - + + + Unlabeled Examples Prob. Learner Prob. Classifier +  +  +  +   +
82 Semi-Supervised EM Training Examples - - + + + Prob. Learner +  +  +  +   + Prob. Classifier Continue retraining iterations until probabilistic labels on unlabeled data converge.
Semi-Supervised Results • Use of additional unlabeled data improves on supervised learning when amount of labeled data is very small and amount of unlabeled data is large. • Can degrade performance when there is sufficient labeled data to learn a decent model and when unsupervised learning tends to create labels that are incompatible with the desired ones. – There are negative results for semi-supervised POS tagging since unsupervised learning tends to learn semantic labels (e.g. eating verbs, animate nouns) that are better at predicting the data than purely syntactic labels (e.g. verb, noun).
Conclusions • POS Tagging is the lowest level of syntactic analysis. • It is an instance of sequence labeling, a collective classification task that also has applications in information extraction, phrase chunking, semantic role labeling, and bioinformatics. • HMMs are a standard generative probabilistic model for sequence labeling that allows for efficiently computing the globally most probable sequence of labels and supports supervised, unsupervised and semi-supervised learning.

More Related Content

PPT
Introduction to Natural Language Processing
Pranav Gupta
 
PDF
Introduction to Natural Language Processing (NLP)
VenkateshMurugadas
 
PDF
Introduction to natural language processing
Minh Pham
 
PPTX
MEBI 591C/598 – Data and Text Mining in Biomedical Informatics
butest
 
PPTX
8_POSNER_university_of Azad_Jammau_kashmir.pptx
mh187782
 
PPTX
Natural Language Processing (NLP).pptx
SHIBDASDUTTA
 
PDF
Natural Language Processing (NLP)
Yuriy Guts
 
PDF
Natural Language Processing with Python
Benjamin Bengfort
 
Introduction to Natural Language Processing
Pranav Gupta
 
Introduction to Natural Language Processing (NLP)
VenkateshMurugadas
 
Introduction to natural language processing
Minh Pham
 
MEBI 591C/598 – Data and Text Mining in Biomedical Informatics
butest
 
8_POSNER_university_of Azad_Jammau_kashmir.pptx
mh187782
 
Natural Language Processing (NLP).pptx
SHIBDASDUTTA
 
Natural Language Processing (NLP)
Yuriy Guts
 
Natural Language Processing with Python
Benjamin Bengfort
 

Similar to Natural Language Processing:Part-Of-Speech Tagging, Sequence Labeling, and Hidden Markov Models (HMMs).ppt (20)

PDF
Crash Course in Natural Language Processing (2016)
Vsevolod Dyomkin
 
PDF
Devday @ Sahaj - Domain Specific NLP Pipelines
Rajesh Muppalla
 
PDF
NLP for Everyday People
Rebecca Bilbro
 
PPTX
lecture 1 intro NLP_lecture 1 intro NLP.pptx
MustansarAli20
 
PPTX
Presentation1
Ritikesh Bhaskarwar
 
PDF
Natural language processing (nlp)
Kuppusamy P
 
PDF
Crash-course in Natural Language Processing
Vsevolod Dyomkin
 
PDF
Natural Language Processing
Geeks Anonymes
 
PPT
Intro 2 document
Uma Kant
 
PDF
9780429149207_previewpdf.pdf
KanwalNaz30
 
PDF
IRJET- A Review on Part-of-Speech Tagging on Gujarati Language
IRJET Journal
 
PPT
NLP Introduction.ppt machine learning presentation
PriyankaRamavath3
 
PPTX
Language models
Maryam Khordad
 
PDF
NLP Project Full Cycle
Vsevolod Dyomkin
 
PPT
slp05.pptnnnnnnnnnnnnnnnnnnnnnnnnnnnnmmmmmmmmm
RAtna29
 
PPTX
Group 5 Text Vectorization in Natural Language Processing.pptx
apsapssingh9
 
PDF
Lausanne 2019 #3
Arthur Charpentier
 
PPTX
Weekairtificial intelligence 8-Module 7 NLP.pptx
funcrusherr
 
PPT
week7.ppt
GiannisPagges
 
PPT
NLTK Python Basic Natural Language Processing.ppt
abdul623429
 
Crash Course in Natural Language Processing (2016)
Vsevolod Dyomkin
 
Devday @ Sahaj - Domain Specific NLP Pipelines
Rajesh Muppalla
 
NLP for Everyday People
Rebecca Bilbro
 
lecture 1 intro NLP_lecture 1 intro NLP.pptx
MustansarAli20
 
Presentation1
Ritikesh Bhaskarwar
 
Natural language processing (nlp)
Kuppusamy P
 
Crash-course in Natural Language Processing
Vsevolod Dyomkin
 
Natural Language Processing
Geeks Anonymes
 
Intro 2 document
Uma Kant
 
9780429149207_previewpdf.pdf
KanwalNaz30
 
IRJET- A Review on Part-of-Speech Tagging on Gujarati Language
IRJET Journal
 
NLP Introduction.ppt machine learning presentation
PriyankaRamavath3
 
Language models
Maryam Khordad
 
NLP Project Full Cycle
Vsevolod Dyomkin
 
slp05.pptnnnnnnnnnnnnnnnnnnnnnnnnnnnnmmmmmmmmm
RAtna29
 
Group 5 Text Vectorization in Natural Language Processing.pptx
apsapssingh9
 
Lausanne 2019 #3
Arthur Charpentier
 
Weekairtificial intelligence 8-Module 7 NLP.pptx
funcrusherr
 
week7.ppt
GiannisPagges
 
NLTK Python Basic Natural Language Processing.ppt
abdul623429
 
Ad

Recently uploaded (20)

PDF
KodekX | Application Modernization Development
KodekX
 
PPT
“AI and Expert System Decision Support & Business Intelligence Systems”
ZubinRadhakrishnan
 
PDF
NewMind AI Weekly Chronicles - August'25 Week I
NewMind AI
 
PDF
Spectral efficient network and resource selection model in 5G networks
IAESIJAI
 
PDF
Electronic commerce courselecture one. Pdf
OmerMohamed64
 
PDF
Blue Purple Modern Animated Computer Science Presentation.pdf.pdf
Akar16
 
PDF
Approach and Philosophy of On baking technology
30anthuong17
 
PDF
Mobile App Security Testing_ A Comprehensive Guide.pdf
flufftailshop
 
PPTX
Digital-Transformation-Roadmap-for-Companies.pptx
David Malick Dieng
 
PDF
Dropbox Q2 2025 Financial Results & Investor Presentation
Dropbox
 
PDF
TokAI - TikTok AI Agent : The First AI Application That Analyzes 10,000+ Vira...
SOFTTECHHUB
 
PDF
Encapsulation theory and applications.pdf
gurumoop
 
PDF
Shreyas Phanse Resume: Experienced Backend Engineer | Java • Spring Boot • Ka...
SHREYAS PHANSE
 
PPTX
KOM of Painting work and Equipment Insulation REV00 update 25-dec.pptx
muhammadmuneebnaeem7
 
PPTX
MYSQL Presentation for SQL database connectivity
Swati270511
 
PPTX
Understanding_Digital_Forensics_Presentation.pptx
ImranKhan423233
 
PPTX
Detection-First SIEM: Rule Types, Dashboards, and Threat-Informed Strategy
ReZa AdineH
 
PPTX
Effective Security Operations Center (SOC) A Modern, Strategic, and Threat-In...
ReZa AdineH
 
PPTX
20250228 LYD VKU AI Blended-Learning.pptx
DatVoNgoc
 
PDF
Agricultural_Statistics_at_a_Glance_2022_0.pdf
SandeepSingh286037
 
KodekX | Application Modernization Development
KodekX
 
“AI and Expert System Decision Support & Business Intelligence Systems”
ZubinRadhakrishnan
 
NewMind AI Weekly Chronicles - August'25 Week I
NewMind AI
 
Spectral efficient network and resource selection model in 5G networks
IAESIJAI
 
Electronic commerce courselecture one. Pdf
OmerMohamed64
 
Blue Purple Modern Animated Computer Science Presentation.pdf.pdf
Akar16
 
Approach and Philosophy of On baking technology
30anthuong17
 
Mobile App Security Testing_ A Comprehensive Guide.pdf
flufftailshop
 
Digital-Transformation-Roadmap-for-Companies.pptx
David Malick Dieng
 
Dropbox Q2 2025 Financial Results & Investor Presentation
Dropbox
 
TokAI - TikTok AI Agent : The First AI Application That Analyzes 10,000+ Vira...
SOFTTECHHUB
 
Encapsulation theory and applications.pdf
gurumoop
 
Shreyas Phanse Resume: Experienced Backend Engineer | Java • Spring Boot • Ka...
SHREYAS PHANSE
 
KOM of Painting work and Equipment Insulation REV00 update 25-dec.pptx
muhammadmuneebnaeem7
 
MYSQL Presentation for SQL database connectivity
Swati270511
 
Understanding_Digital_Forensics_Presentation.pptx
ImranKhan423233
 
Detection-First SIEM: Rule Types, Dashboards, and Threat-Informed Strategy
ReZa AdineH
 
Effective Security Operations Center (SOC) A Modern, Strategic, and Threat-In...
ReZa AdineH
 
20250228 LYD VKU AI Blended-Learning.pptx
DatVoNgoc
 
Agricultural_Statistics_at_a_Glance_2022_0.pdf
SandeepSingh286037
 
Ad

Natural Language Processing:Part-Of-Speech Tagging, Sequence Labeling, and Hidden Markov Models (HMMs).ppt

  • 1. 1 1 CS 388: Natural Language Processing: Part-Of-Speech Tagging, Sequence Labeling, and Hidden Markov Models (HMMs) Raymond J. Mooney University of Texas at Austin
  • 2. 2 Part Of Speech Tagging • Annotate each word in a sentence with a part-of-speech marker. • Lowest level of syntactic analysis. • Useful for subsequent syntactic parsing and word sense disambiguation. John saw the saw and decided to take it to the table. NNP VBD DT NN CC VBD TO VB PRP IN DT NN
  • 3. 3 English POS Tagsets • Original Brown corpus used a large set of 87 POS tags. • Most common in NLP today is the Penn Treebank set of 45 tags. – Tagset used in these slides. – Reduced from the Brown set for use in the context of a parsed corpus (i.e. treebank). • The C5 tagset used for the British National Corpus (BNC) has 61 tags.
  • 4. 4 English Parts of Speech • Noun (person, place or thing) – Singular (NN): dog, fork – Plural (NNS): dogs, forks – Proper (NNP, NNPS): John, Springfields – Personal pronoun (PRP): I, you, he, she, it – Wh-pronoun (WP): who, what • Verb (actions and processes) – Base, infinitive (VB): eat – Past tense (VBD): ate – Gerund (VBG): eating – Past participle (VBN): eaten – Non 3rd person singular present tense (VBP): eat – 3rd person singular present tense: (VBZ): eats – Modal (MD): should, can – To (TO): to (to eat)
  • 5. 5 English Parts of Speech (cont.) • Adjective (modify nouns) – Basic (JJ): red, tall – Comparative (JJR): redder, taller – Superlative (JJS): reddest, tallest • Adverb (modify verbs) – Basic (RB): quickly – Comparative (RBR): quicker – Superlative (RBS): quickest • Preposition (IN): on, in, by, to, with • Determiner: – Basic (DT) a, an, the – WH-determiner (WDT): which, that • Coordinating Conjunction (CC): and, but, or, • Particle (RP): off (took off), up (put up)
  • 6. Closed vs. Open Class • Closed class categories are composed of a small, fixed set of grammatical function words for a given language. – Pronouns, Prepositions, Modals, Determiners, Particles, Conjunctions • Open class categories have large number of words and new ones are easily invented. – Nouns (Googler, textlish), Verbs (Google), Adjectives (geeky), Abverb (automagically) 6
  • 7. 7 Ambiguity in POS Tagging • “Like” can be a verb or a preposition – I like/VBP candy. – Time flies like/IN an arrow. • “Around” can be a preposition, particle, or adverb – I bought it at the shop around/IN the corner. – I never got around/RP to getting a car. – A new Prius costs around/RB $25K.
  • 8. 8 POS Tagging Process • Usually assume a separate initial tokenization process that separates and/or disambiguates punctuation, including detecting sentence boundaries. • Degree of ambiguity in English (based on Brown corpus) – 11.5% of word types are ambiguous. – 40% of word tokens are ambiguous. • Average POS tagging disagreement amongst expert human judges for the Penn treebank was 3.5% – Based on correcting the output of an initial automated tagger, which was deemed to be more accurate than tagging from scratch. • Baseline: Picking the most frequent tag for each specific word type gives about 90% accuracy – 93.7% if use model for unknown words for Penn Treebank tagset.
  • 9. 9 POS Tagging Approaches • Rule-Based: Human crafted rules based on lexical and other linguistic knowledge. • Learning-Based: Trained on human annotated corpora like the Penn Treebank. – Statistical models: Hidden Markov Model (HMM), Maximum Entropy Markov Model (MEMM), Conditional Random Field (CRF) – Rule learning: Transformation Based Learning (TBL) – Neural networks: Recurrent networks like Long Short Term Memory (LSTMs) • Generally, learning-based approaches have been found to be more effective overall, taking into account the total amount of human expertise and effort involved.
  • 10. 10 Classification Learning • Typical machine learning addresses the problem of classifying a feature-vector description into a fixed number of classes. • There are many standard learning methods for this task: – Decision Trees and Rule Learning – Naïve Bayes and Bayesian Networks – Logistic Regression / Maximum Entropy (MaxEnt) – Perceptron and Neural Networks – Support Vector Machines (SVMs) – Nearest-Neighbor / Instance-Based
  • 11. 11 Beyond Classification Learning • Standard classification problem assumes individual cases are disconnected and independent (i.i.d.: independently and identically distributed). • Many NLP problems do not satisfy this assumption and involve making many connected decisions, each resolving a different ambiguity, but which are mutually dependent. • More sophisticated learning and inference techniques are needed to handle such situations in general.
  • 12. 12 Sequence Labeling Problem • Many NLP problems can viewed as sequence labeling. • Each token in a sequence is assigned a label. • Labels of tokens are dependent on the labels of other tokens in the sequence, particularly their neighbors (not i.i.d). foo bar blam zonk zonk bar blam
  • 13. 13 Information Extraction • Identify phrases in language that refer to specific types of entities and relations in text. • Named entity recognition is task of identifying names of people, places, organizations, etc. in text. people organizations places – Michael Dell is the CEO of Dell Computer Corporation and lives in Austin Texas. • Extract pieces of information relevant to a specific application, e.g. used car ads: make model year mileage price – For sale, 2002 Toyota Prius, 20,000 mi, $15K or best offer. Available starting July 30, 2006.
  • 14. 14 Semantic Role Labeling • For each clause, determine the semantic role played by each noun phrase that is an argument to the verb. agent patient source destination instrument – John drove Mary from Austin to Dallas in his Toyota Prius. – The hammer broke the window. • Also referred to a “case role analysis,” “thematic analysis,” and “shallow semantic parsing”
  • 15. 15 Bioinformatics • Sequence labeling also valuable in labeling genetic sequences in genome analysis. extron intron – AGCTAACGTTCGATACGGATTACAGCCT
  • 16. 16 Problems with Sequence Labeling as Classification • Not easy to integrate information from category of tokens on both sides. • Difficult to propagate uncertainty between decisions and “collectively” determine the most likely joint assignment of categories to all of the tokens in a sequence.
  • 17. 17 Probabilistic Sequence Models • Probabilistic sequence models allow integrating uncertainty over multiple, interdependent classifications and collectively determine the most likely global assignment. • Two standard models – Hidden Markov Model (HMM) – Conditional Random Field (CRF)
  • 18. 18 Markov Model / Markov Chain • A finite state machine with probabilistic state transitions. • Makes Markov assumption that next state only depends on the current state and independent of previous history.
  • 19. 19 Sample Markov Model for POS 0.95 0.9 0.05 stop 0.5 0.1 0.8 0.1 0.1 0.25 0.25 start 0.1 0.5 0.4 Det Noun PropNoun Verb
  • 20. 20 Sample Markov Model for POS 0.95 0.9 0.05 stop 0.5 0.1 0.8 0.1 0.1 0.25 0.25 start 0.1 0.5 0.4 Det Noun PropNoun Verb P(PropNoun Verb Det Noun) = 0.4*0.8*0.25*0.95*0.1=0.0076
  • 21. 21 Hidden Markov Model • Probabilistic generative model for sequences. • Assume an underlying set of hidden (unobserved, latent) states in which the model can be (e.g. parts of speech). • Assume probabilistic transitions between states over time (e.g. transition from POS to another POS as sequence is generated). • Assume a probabilistic generation of tokens from states (e.g. words generated for each POS).
  • 22. 22 Sample HMM for POS PropNoun JohnMary Alice Jerry Tom Noun cat dog car pen bed apple Det a the th e the that a the a Verb bit ate saw played hit 0.95 0.9 gave 0.05 stop 0.5 0.1 0.8 0.1 0.1 0.25 0.25 start 0.1 0.5 0.4
  • 23. 23 Sample HMM Generation PropNoun JohnMary Alice Jerry Tom Noun cat dog car pen bed apple Det a the th e the that a the a Verb bit ate saw played hit 0.95 0.9 gave 0.05 stop 0.5 0.1 0.8 0.1 0.1 0.25 0.25 start 0.1 0.5 0.4
  • 24. 24 Sample HMM Generation PropNoun JohnMary Alice Jerry Tom Noun cat dog car pen bed apple Det a the th e the that a the a Verb bit ate saw played hit 0.95 0.9 gave 0.05 stop 0.5 0.1 0.8 0.1 0.1 0.25 start 0.1 0.5 0.4
  • 25. 25 Sample HMM Generation PropNoun JohnMary Alice Jerry Tom Noun cat dog car pen bed apple Det a the th e the that a the a Verb bit ate saw played hit 0.95 0.9 gave 0.05 stop 0.5 0.1 0.8 0.1 0.1 0.25 0.25 John start 0.1 0.5 0.4
  • 26. 26 Sample HMM Generation PropNoun JohnMary Alice Jerry Tom Noun cat dog car pen bed apple Det a the th e the that a the a Verb bit ate saw played hit 0.95 0.9 gave 0.05 stop 0.5 0.1 0.8 0.1 0.1 0.25 0.25 John start 0.1 0.5 0.4
  • 27. 27 Sample HMM Generation PropNoun JohnMary Alice Jerry Tom Noun cat dog car pen bed apple Det a the th e the that a the a Verb bit ate saw played hit 0.95 0.9 gave 0.05 stop 0.5 0.1 0.8 0.1 0.1 0.25 0.25 John bit start 0.1 0.5 0.4
  • 28. 28 Sample HMM Generation PropNoun JohnMary Alice Jerry Tom Noun cat dog car pen bed apple Det a the th e the that a the a Verb bit ate saw played hit 0.95 0.9 gave 0.05 stop 0.5 0.1 0.8 0.1 0.1 0.25 0.25 John bit start 0.1 0.5 0.4
  • 29. 29 Sample HMM Generation PropNoun JohnMary Alice Jerry Tom Noun cat dog car pen bed apple Det a the th e the that a the a Verb bit ate saw played hit 0.95 0.9 gave 0.05 stop 0.5 0.1 0.8 0.1 0.1 0.25 0.25 John bit the start 0.1 0.5 0.4
  • 30. 30 Sample HMM Generation PropNoun JohnMary Alice Jerry Tom Noun cat dog car pen bed apple Det a the th e the that a the a Verb bit ate saw played hit 0.95 0.9 gave 0.05 stop 0.5 0.1 0.8 0.1 0.1 0.25 0.25 John bit the start 0.1 0.5 0.4
  • 31. 31 Sample HMM Generation PropNoun JohnMary Alice Jerry Tom Noun cat dog car pen bed apple Det a the th e the that a the a Verb bit ate saw played hit 0.95 0.9 gave 0.05 stop 0.5 0.1 0.8 0.1 0.1 0.25 0.25 John bit the apple start 0.1 0.5 0.4
  • 32. 32 Sample HMM Generation PropNoun JohnMary Alice Jerry Tom Noun cat dog car pen bed apple Det a the th e the that a the a Verb bit ate saw played hit 0.95 0.9 gave 0.05 stop 0.5 0.1 0.8 0.1 0.1 0.25 0.25 John bit the apple start 0.1 0.5 0.4
  • 33. 33 Formal Definition of an HMM • A set of N +2 states S={s0,s1,s2, … sN, sF} – Distinguished start state: s0 – Distinguished final state: sF • A set of M possible observations V={v1,v2…vM} • A state transition probability distribution A={aij} • Observation probability distribution for each state j B={bj(k)} • Total parameter set λ={A,B} F j i N j i s q s q P a i t j t ij         , 0 and , 1 ) | ( 1 M k 1 1 ) | at ( ) (       N j s q t v P k b j t k j N i a a iF N j ij       0 1 1
  • 34. 34 HMM Generation Procedure • To generate a sequence of T observations: O = o1 o2 … oT Set initial state q1=s0 For t = 1 to T Transit to another state qt+1=sj based on transition distribution aij for state qt Pick an observation ot=vk based on being in state qt using distribution bqt(k)
  • 35. 35 Three Useful HMM Tasks • Observation Likelihood: To classify and order sequences. • Most likely state sequence (Decoding): To tag each token in a sequence with a label. • Maximum likelihood training (Learning): To train models to fit empirical training data.
  • 36. 36 HMM: Observation Likelihood • Given a sequence of observations, O, and a model with a set of parameters, λ, what is the probability that this observation was generated by this model: P(O| λ) ? • Allows HMM to be used as a language model: A formal probabilistic model of a language that assigns a probability to each string saying how likely that string was to have been generated by the language. • Useful for two tasks: – Sequence Classification – Most Likely Sequence
  • 37. 37 Sequence Classification • Assume an HMM is available for each category (i.e. language). • What is the most likely category for a given observation sequence, i.e. which category’s HMM is most likely to have generated it? • Used in speech recognition to find most likely word model to have generate a given sound or phoneme sequence. Austin Boston ? ? P(O | Austin) > P(O | Boston) ? ah s t e n O
  • 38. 38 Most Likely Sequence • Of two or more possible sequences, which one was most likely generated by a given model? • Used to score alternative word sequence interpretations in speech recognition. Ordinary English dice precedent core vice president Gore O1 O2 ? ? P(O2 | OrdEnglish) > P(O1 | OrdEnglish) ?
  • 39. 39 HMM: Observation Likelihood Naïve Solution • Consider all possible state sequences, Q, of length T that the model could have traversed in generating the given observation sequence. • Compute the probability of a given state sequence from A, and multiply it by the probabilities of generating each of given observations in each of the corresponding states in this sequence to get P(O,Q| λ) = P(O| Q, λ) P(Q| λ) . • Sum this over all possible state sequences to get P(O| λ). • Computationally complex: O(TNT ).
  • 40. 40 HMM: Observation Likelihood Efficient Solution • Due to the Markov assumption, the probability of being in any state at any given time t only relies on the probability of being in each of the possible states at time t−1. • Forward Algorithm: Uses dynamic programming to exploit this fact to efficiently compute observation likelihood in O(TN2 ) time. – Compute a forward trellis that compactly and implicitly encodes information about all possible state paths.
  • 41. Forward Trellis 41 s1 s2 sN       s0 sF                      t1 t2 t3 tT-1 tT • Continue forward in time until reaching final time point and sum probability of ending in final state.
  • 42. Forward Probabilities • Let t(j) be the probability of being in state j after seeing the first t observations (by summing over all initial paths leading to j). 42 ) | , ,... , ( ) ( 2 1   j t t t s q o o o P j  
  • 43. Forward Step 43 s1 s2 sN    sj t-1(i) t(i) a1j a2j aNj a2j • Consider all possible ways of getting to sj at time t by coming from all possible states si and determine probability of each. • Sum these to get the total probability of being in state sj at time t while accounting for the first t −1 observations. • Then multiply by the probability of actually observing ot in sj.
  • 44. Computing the Forward Probabilities • Initialization • Recursion • Termination 44 N j o b a j j j    1 ) ( ) ( 1 0 1  T t N j o b a i j t j N i ij t t               1 , 1 ) ( ) ( ) ( 1 1        N i iF T F T a i s O P 1 1 ) ( ) ( ) | (   
  • 45. Forward Computational Complexity • Requires only O(TN2 ) time to compute the probability of an observed sequence given a model. • Exploits the fact that all state sequences must merge into one of the N possible states at any point in time and the Markov assumption that only the last state effects the next one. 45
  • 46. 46 Most Likely State Sequence (Decoding) • Given an observation sequence, O, and a model, λ, what is the most likely state sequence,Q=q1,q2,…qT, that generated this sequence from this model? • Used for sequence labeling, assuming each state corresponds to a tag, it determines the globally best assignment of tags to all tokens in a sequence using a principled approach grounded in probability theory. John gave the dog an apple.
  • 47. 47 Most Likely State Sequence • Given an observation sequence, O, and a model, λ, what is the most likely state sequence,Q=q1,q2,…qT, that generated this sequence from this model? • Used for sequence labeling, assuming each state corresponds to a tag, it determines the globally best assignment of tags to all tokens in a sequence using a principled approach grounded in probability theory. John gave the dog an apple. Det Noun PropNoun Verb
  • 48. 48 Most Likely State Sequence • Given an observation sequence, O, and a model, λ, what is the most likely state sequence,Q=q1,q2,…qT, that generated this sequence from this model? • Used for sequence labeling, assuming each state corresponds to a tag, it determines the globally best assignment of tags to all tokens in a sequence using a principled approach grounded in probability theory. John gave the dog an apple. Det Noun PropNoun Verb
  • 49. 49 Most Likely State Sequence • Given an observation sequence, O, and a model, λ, what is the most likely state sequence,Q=q1,q2,…qT, that generated this sequence from this model? • Used for sequence labeling, assuming each state corresponds to a tag, it determines the globally best assignment of tags to all tokens in a sequence using a principled approach grounded in probability theory. John gave the dog an apple. Det Noun PropNoun Verb
  • 50. 50 Most Likely State Sequence • Given an observation sequence, O, and a model, λ, what is the most likely state sequence,Q=q1,q2,…qT, that generated this sequence from this model? • Used for sequence labeling, assuming each state corresponds to a tag, it determines the globally best assignment of tags to all tokens in a sequence using a principled approach grounded in probability theory. John gave the dog an apple. Det Noun PropNoun Verb
  • 51. 51 Most Likely State Sequence • Given an observation sequence, O, and a model, λ, what is the most likely state sequence,Q=q1,q2,…qT, that generated this sequence from this model? • Used for sequence labeling, assuming each state corresponds to a tag, it determines the globally best assignment of tags to all tokens in a sequence using a principled approach grounded in probability theory. John gave the dog an apple. Det Noun PropNoun Verb
  • 52. 52 Most Likely State Sequence • Given an observation sequence, O, and a model, λ, what is the most likely state sequence,Q=q1,q2,…qT, that generated this sequence from this model? • Used for sequence labeling, assuming each state corresponds to a tag, it determines the globally best assignment of tags to all tokens in a sequence using a principled approach grounded in probability theory. John gave the dog an apple. Det Noun PropNoun Verb
  • 53. 53 HMM: Most Likely State Sequence Efficient Solution • Obviously, could use naïve algorithm based on examining every possible state sequence of length T. • Dynamic Programming can also be used to exploit the Markov assumption and efficiently determine the most likely state sequence for a given observation and model. • Standard procedure is called the Viterbi algorithm (Viterbi, 1967) and also has O(N2 T) time complexity.
  • 54. Viterbi Scores • Recursively compute the probability of the most likely subsequence of states that accounts for the first t observations and ends in state sj. 54 ) | , ,..., , ,..., , ( max ) ( 1 1 1 0 ,..., , 1 1 0  j t t t q q q t s q o o q q q P j v t     • Also record “backpointers” that subsequently allow backtracing the most probable state sequence.  btt(j) stores the state at time t-1 that maximizes the probability that system was in state sj at time t (given the observed sequence).
  • 55. Computing the Viterbi Scores • Initialization • Recursion • Termination 55 N j o b a j v j j    1 ) ( ) ( 1 0 1 T t N j o b a i v j v t j ij t N i t        1 , 1 ) ( ) ( max ) ( 1 1 iF T N i F T a i v s v P ) ( max ) ( * 1 1     Analogous to Forward algorithm except take max instead of sum
  • 56. Computing the Viterbi Backpointers • Initialization • Recursion • Termination 56 N j s j bt    1 ) ( 0 1 T t N j o b a i v j bt t j ij t N i t        1 , 1 ) ( ) ( argmax ) ( 1 1 iF T N i F T T a i v s bt q ) ( argmax ) ( * 1 1     Final state in the most probable state sequence. Follow backpointers to initial state to construct full sequence.
  • 58. Viterbi Backtrace 58 s1 s2 sN       s0 sF                      t1 t2 t3 tT-1 tT Most likely Sequence: s0 sN s1 s2 …s2 sF
  • 59. HMM Learning • Supervised Learning: All training sequences are completely labeled (tagged). • Unsupervised Learning: All training sequences are unlabelled (but generally know the number of tags, i.e. states). • Semisupervised Learning: Some training sequences are labeled, most are unlabeled. 59
  • 60. 60 Supervised HMM Training • If training sequences are labeled (tagged) with the underlying state sequences that generated them, then the parameters, λ={A,B} can all be estimated directly. Supervised HMM Training John ate the apple A dog bit Mary Mary hit the dog John gave Mary the cat. . . . Training Sequences Det Noun PropNoun Verb
  • 61. Supervised Parameter Estimation • Estimate state transition probabilities based on tag bigram and unigram statistics in the labeled data. • Estimate the observation probabilities based on tag/word co-occurrence statistics in the labeled data. • Use appropriate smoothing if training data is sparse. 61 ) ( ) q , ( 1 t i t j i t ij s q C s s q C a      ) ( ) , ( ) ( j i k i j i j s q C v o s q C k b    
  • 62. Learning and Using HMM Taggers • Use a corpus of labeled sequence data to easily construct an HMM using supervised training. • Given a novel unlabeled test sequence to tag, use the Viterbi algorithm to predict the most likely (globally optimal) tag sequence. 62
  • 63. Evaluating Taggers • Train on training set of labeled sequences. • Possibly tune parameters based on performance on a development set. • Measure accuracy on a disjoint test set. • Generally measure tagging accuracy, i.e. the percentage of tokens tagged correctly. • Accuracy of most modern POS taggers, including HMMs is 96−97% (for Penn tagset trained on about 800K words) . – Generally matching human agreement level. 63
  • 64. 64 Unsupervised Maximum Likelihood Training ah s t e n a s t i n oh s t u n eh z t en . . . Training Sequences HMM Training Austin
  • 65. 65 Maximum Likelihood Training • Given an observation sequence, O, what set of parameters, λ, for a given model maximizes the probability that this data was generated from this model (P(O| λ))? • Used to train an HMM model and properly induce its parameters from a set of training data. • Only need to have an unannotated observation sequence (or set of sequences) generated from the model. Does not need to know the correct state sequence(s) for the observation sequence(s). In this sense, it is unsupervised.
  • 66. Bayes Theorem Simple proof from definition of conditional probability: ) ( ) ( ) | ( ) | ( E P H P H E P E H P  ) ( ) ( ) | ( E P E H P E H P   ) ( ) ( ) | ( H P E H P H E P   ) ( ) | ( ) ( H P H E P E H P   QED: (Def. cond. prob.) (Def. cond. prob.) ) ( ) ( ) | ( ) | ( E P H P H E P E H P 
  • 67. Maximum Likelihood vs. Maximum A Posteriori (MAP) • The MAP parameter estimate is the most likely given the observed data, O. ) ( ) ( ) | ( argmax ) | ( argmax O P P O P O P MAP         ) ( ) | ( argmax    P O P • If all parameterizations are assumed to be equally likely a priori, then MLE and MAP are the same. • If parameters are given priors (e.g. Gaussian or Lapacian with zero mean), then MAP is a principled way to perform smoothing or regularization.
  • 68. 68 HMM: Maximum Likelihood Training Efficient Solution • There is no known efficient algorithm for finding the parameters, λ, that truly maximizes P(O| λ). • However, using iterative re-estimation, the Baum- Welch algorithm (a.k.a. forward-backward) , a version of a standard statistical procedure called Expectation Maximization (EM), is able to locally maximize P(O| λ). • In practice, EM is able to find a good set of parameters that provide a good fit to the training data in many cases.
  • 69. 69 EM Algorithm • Iterative method for learning probabilistic categorization model from unsupervised data. • Initially assume random assignment of examples to categories. • Learn an initial probabilistic model by estimating model parameters  from this randomly labeled data. • Iterate following two steps until convergence: – Expectation (E-step): Compute P(ci | E) for each example given the current model, and probabilistically re-label the examples based on these posterior probability estimates. – Maximization (M-step): Re-estimate the model parameters, , from the probabilistically re-labeled data.
  • 70. 70 EM Unlabeled Examples +  +  +  +   + Assign random probabilistic labels to unlabeled data Initialize:
  • 71. 71 EM Prob. Learner +  +  +  +   + Give soft-labeled training data to a probabilistic learner Initialize:
  • 72. 72 EM Prob. Learner Prob. Classifier +  +  +  +   + Produce a probabilistic classifier Initialize:
  • 73. 73 EM Prob. Learner Prob. Classifier Relabel unlabled data using the trained classifier +  +  +  +   + E Step:
  • 74. 74 EM Prob. Learner +  +  +  +   + Prob. Classifier Continue EM iterations until probabilistic labels on unlabeled data converge. Retrain classifier on relabeled data M step:
  • 75. 75 Sketch of Baum-Welch (EM) Algorithm for Training HMMs Assume an HMM with N states. Randomly set its parameters λ=(A,B) (making sure they represent legal distributions) Until converge (i.e. λ no longer changes) do: E Step: Use the forward/backward procedure to determine the probability of various possible state sequences for generating the training data M Step: Use these probability estimates to re-estimate values for all of the parameters λ See textbook for detailed equations for E and M steps
  • 76. EM Properties • Each iteration changes the parameters in a way that is guaranteed to increase the likelihood of the data: P(O|). • Anytime algorithm: Can stop at any time prior to convergence to get approximate solution. • Converges to a local maximum.
  • 77. Semi-Supervised Learning • EM algorithms can be trained with a mix of labeled and unlabeled data. • EM basically predicts a probabilistic (soft) labeling of the instances and then iteratively retrains using supervised learning on these predicted labels (“self training”). • EM can also exploit supervised data: – 1) Use supervised learning on labeled data to initialize the parameters (instead of initializing them randomly). – 2) Use known labels for supervised data instead of predicting soft labels for these examples during retraining iterations.
  • 78. 78 Semi-Supervised EM Training Examples - - + + + Unlabeled Examples Prob. Learner Prob. Classifier +  +  +  +   +
  • 79. 79 Semi-Supervised EM Training Examples - - + + + Prob. Learner +  +  +  +   + Prob. Classifier
  • 80. 80 Semi-Supervised EM Training Examples - - + + + Prob. Learner +  +  +  +   + Prob. Classifier
  • 81. 81 Semi-Supervised EM Training Examples - - + + + Unlabeled Examples Prob. Learner Prob. Classifier +  +  +  +   +
  • 82. 82 Semi-Supervised EM Training Examples - - + + + Prob. Learner +  +  +  +   + Prob. Classifier Continue retraining iterations until probabilistic labels on unlabeled data converge.
  • 83. Semi-Supervised Results • Use of additional unlabeled data improves on supervised learning when amount of labeled data is very small and amount of unlabeled data is large. • Can degrade performance when there is sufficient labeled data to learn a decent model and when unsupervised learning tends to create labels that are incompatible with the desired ones. – There are negative results for semi-supervised POS tagging since unsupervised learning tends to learn semantic labels (e.g. eating verbs, animate nouns) that are better at predicting the data than purely syntactic labels (e.g. verb, noun).
  • 84. Conclusions • POS Tagging is the lowest level of syntactic analysis. • It is an instance of sequence labeling, a collective classification task that also has applications in information extraction, phrase chunking, semantic role labeling, and bioinformatics. • HMMs are a standard generative probabilistic model for sequence labeling that allows for efficiently computing the globally most probable sequence of labels and supports supervised, unsupervised and semi-supervised learning.