Semi-Supervised Learning over Text

1. Semi-Supervised Learning over Text Tom M. Mitchell Machine Learning Department Carnegie Mellon University September 2006

2. Statistical learning methods require LOTS of training data Can we use all that unlabelled text?

3. Outline • Maximizing likelihood in probabilistic models • EM for text classification • Co-Training and redundantly predictive features • Document classification • Named entity recognition • Theoretical analysis • Sample of additional tasks • Word sense disambiguation • Learning HTML-based extractors • Large-scale bootstrapping: extracting from the web

4. Many text learning tasks • Document classification. • f: Doc  Class • Spam filtering, relevance rating, web page classification, ... • and unsupervised document clustering • Information extraction. • f: Sentence  Fact, f: Doc  Facts • Parsing • f: Sentence  ParseTree • Related: part-of-speech tagging, co-reference res., prep phrase attachment • Translation • f: EnglishDoc  FrenchDoc

5. Semi-supervised Document classification (probabilistic model and EM)

6. Document Classification: Bag of Words Approach aardvark 0 about 2 all 2 Africa 1 apple 0 anxious 0 ... gas 1 ... oil 1 … Zaire 0

7. Accuracy vs. # training examples For code and data, seewww.cs.cmu.edu/~tom/mlbook.html click on “Software and Data”

8. Y X1 X2 X3 X4 What if we have labels for only some documents? Learn P(Y|X) • EM: Repeat until convergence • Use probabilistic labels to train classifier h • Apply h to assign probabilistic labels to unlabeled data

9. From [Nigam et al., 2000]

10. E Step: wt is t-th word in vocabulary M Step:

11. Using one labeled example per class Words sorted by P(w|course) / P(w| : course)

12. 20 Newsgroups

13. 20 Newsgroups

14. Elaboration 1: Downweight the influence of unlabeled examples by factor l Chosen by cross validation New M step:

15. Why/When will this work? • What’s best case? Worst case? How can we test which we have?

16. EM for Semi-Supervised Doc Classification • If all data is labeled, corresponds to supervised training of Naïve Bayes classifier • If all data unlabeled, corresponds to mixture-of-multinomial clustering • If both labeled and unlabeled data, it helps if and only if the mixture-of-multinomial modeling assumption is correct • Of course we could extend this to Bayes net models other than Naïve Bayes (e.g., TAN tree) • Other extensions: model negative class as mixture of N multinomials

17. 2. Using Redundantly Predictive Features (Co-Training)

18. Redundantly Predictive Features my advisor Professor Faloutsos

19. Co-Training Key idea: Classifier1and ClassifierJ must: 1. Correctly classify labeled examples 2. Agree on classification of unlabeled Answer1 Answer2 Classifier1 Classifier2

20. CoTraining Algorithm #1 [Blum&Mitchell, 1998] • Given: labeled data L, • unlabeled data U • Loop: • Train g1 (hyperlink classifier) using L • Train g2 (page classifier) using L • Allow g1 to label p positive, n negative examps from U • Allow g2 to label p positive, n negative examps from U • Add these self-labeled examples to L

21. CoTraining: Experimental Results • begin with 12 labeled web pages (academic course) • provide 1,000 additional unlabeled web pages • average error: learning from labeled data 11.1%; • average error: cotraining 5.0% Typical run:

22. Co-Training for Named Entity Extraction(i.e.,classifying which strings refer to people, places, dates, etc.) [Riloff&Jones 98; Collins et al., 98; Jones 05] Answer1 Answer2 Classifier1 Classifier2 New York I flew to ____ today I flew to New York today.

23. CoTraining setting: • wish to learn f: X  Y, given L and U drawn from P(X) • features describing X can be partitioned (X = X1 x X2) • such that f can be computed from either X1 or X2 One result [Blum&Mitchell 1998]: • If • X1 and X2 are conditionally independent given Y • f is PAC learnable from noisy labeled data • Then • f is PAC learnable from weak initial classifier plus unlabeled data

24. pages hyperlinks My advisor + - - Co-Training Rote Learner + + + + + + + + - - - - - - - - - -

25. Co Training • What’s the best-case graph? (most benefit from unlabeled data) • What the worst case? • What does conditional-independence imply about graph? - + + x1 x2

26. Where g is the jth connected component of graph of L+U, m is number of labeled examples j Expected Rote CoTraining error given m examples

27. How many unlabeled examples suffice? Want to assure that connected components in the underlying distribution, GD, are connected components in the observed sample, GS GD GS O(log(N)/) examples assure that with high probability, GS has same connected components as GD [Karger, 94] N is size of GD,  is min cut over all connected components of GD

28. PAC Generalization Bounds on CoTraining [Dasgupta et al., NIPS 2001] This theorem assumes X1 and X2 are conditionally independent given Y

29. Co-Training Theory How can we tune learning environment to enhance effectiveness of Co-Training? # labeled examples # Redundantly predictive inputs # unlabeled examples dependencies among input features Final Accuracy Correctness of confidence assessments  best: inputs conditionally indep given class, increased number of redundant inputs, …

30. + + - + What if CoTraining Assumption Not Perfectly Satisfied? • Idea: Want classifiers that produce a maximally consistent labeling of the data • If learning is an optimization problem, what function should we optimize?

31. Example 2: Learning to extract named entities location? I arrived in Beijing on Saturday. If: “I arrived in <X> on Saturday.” Then: Location(X)

32. Co-Training for Named Entity Extraction(i.e.,classifying which strings refer to people, places, dates, etc.) [Riloff&Jones 98; Collins et al., 98; Jones 05] Answer1 Answer2 Classifier1 Classifier2 Beijing I arrived in __ saturday I arrived in Beijing saturday.

33. Bootstrap learning to extract named entities[Riloff and Jones, 1999], [Collins and Singer, 1999], ... South Africa United Kingdom Warrenton Far_East Oregon Lexington Europe U.S._A. Eastern Canada Blair Southwestern_states Texas States Singapore … Thailand Maine production_control northern_Los New_Zealand eastern_Europe Americas Michigan New_Hampshire Hungary south_america district Latin_America Florida ... Initialization Australia Canada China England France Germany Japan Mexico Switzerland United_states … ... Iterations locations in ?x operations in ?x republic of ?x

34. Co-EM [Nigam & Ghani, 2000; Jones 2005] Idea: • Like co-training, use one set of features to label the other • Like EM, iterate • Assigning probabilistic values to unobserved class labels • Updating model parameters (= labels of other feature set)

35. CoEM applied to Named Entity Recognition[Rosie Jones, 2005], [Ghani & Nigam, 2000] Update rules:

36. CoEM applied to Named Entity Recognition[Rosie Jones, 2005], [Ghani & Nigam, 2000] Update rules:

37. CoEM applied to Named Entity Recognition[Rosie Jones, 2005], [Ghani & Nigam, 2000] Update rules:

39. [Jones, 2005] Can use this for active learning...

40. [Jones, 2005]

41. + + - + What if CoTraining Assumption Not Perfectly Satisfied? • Idea: Want classifiers that produce a maximally consistent labeling of the data • If learning is an optimization problem, what function should we optimize?

42. Error on labeled examples Disagreement over unlabeled What Objective Function? Misfit to estimated class priors

43. What Function Approximators? • Same functional form as logistic regression • Use gradient descent to simultaneously learn g1 and g2, directly minimizing E = E1 + E2 + E3 + E4 • No word independence assumption, use both labeled and unlabeled data

44. .15 .11 * * Gradient CoTrainingClassifying Capitalized sequences as Person Names Eg., “Company presidentMary Smithsaid today…” x1 x2 x1 Error Rates 25 labeled 5000 unlabeled 2300 labeled 5000 unlabeled Using labeled data only .13 .24 Cotraining Cotraining without fitting class priors (E4) * .27 * sensitive to weights of error terms E3 and E4

45. Example 3: Word sense disambiguation [Yarowsky] • “bank” = river bank, or financial bank?? • Assumes a single word sense per document • X1: the document containing the word • X2: the immediate context of the word (‘swim near the __’) Successfully learns “context  word sense” rules when word occurs multiples times in documents.

46. Example 4: Bootstrap learning for IE from HTML structure[Muslea, et al. 2001] X1: HTML preceding the target X2: HTML following the target

47. Example Bootstrap learning algorithms: • Classifying web pages [Blum&Mitchell 98; Slattery 99] • Classifying email [Kiritchenko&Matwin 01; Chan et al. 04] • Named entity extraction [Collins&Singer 99; Jones&Riloff 99] • Wrapper induction [Muslea et al., 01; Mohapatra et al. 04] • Word sense disambiguation [Yarowsky 96] • Discovering new word senses [Pantel&Lin 02] • Synonym discovery [Lin et al., 03] • Relation extraction [Brin et al.; Yangarber et al. 00] • Statistical parsing [Sarkar 01]

48. What to Know • Several approaches to semi-supervised learning • EM with probabilistic model • Co-Training • Graph similarity methods • ... • See reading list below • Redundancy is important • Much more to be done: • Better theoretical models of when/how unlabeled data can help • Bootstrap learning from the web (e.g. Etzioni, 2005, 2006) • Active learning (use limited labeling time of human wisely) • Never ending bootstrap learning? • ...

49. Further Reading • Semi-Supervised Learning, Olivier Chapelle, Bernhard Sch¨olkopf, and Alexander Zien (eds.), MIT Press, 2006. • Semi-Supervised Learning Literature Survey, Xiaojin Zhu, 2006. • Unsupervised word sense disambiguation rivaling supervised methods D. Yarowsky (1995) • "Semi-Supervised Text Classification Using EM,"  K. Nigam, A. McCallum, and T. Mitchell, in Semi-Supervised Learning, Olivier Chapelle, Bernhard Sch¨olkopf, and Alexander Zien (eds.), MIT Press, 2006. • " Text Classification from Labeled and Unlabeled Documents using EM," K. Nigam, Andrew McCallum, Sebastian Thrun and Tom Mitchell. Machine Learning, Kluwer Academic Press, 1999. • " Combining Labeled and Unlabeled Data with Co-Training," A. Blum and T. Mitchell, Proceedings of the 1998 Conference on Computational Learning Theory, July 1998. • Discovering Word Senses from Text Pantel & Lin (2002) • Creating Subjective and Objective Sentence Classifiers from Unannotated Texts by Janyce Wiebe and Ellen Riloff (2005) • Graph Based Semi-Supervised Approach for Information Extraction by Hany Hassan, Ahmed Hassan and Sara Noeman (2006) • The use of unlabeled data to improve supervised learning for text summarization by MR Amini, P Gallinari (2002)

50. Further Reading • Yusuke Shinyama and Satoshi Sekine. Preemptive Information Extraction using Unrestricted Relation Discovery • Alexandre Klementiev and Dan Roth. Named Entity Transliteration and Discovery from Multilingual Comparable Corpora. • Rion L. Snow, Daniel Jurafsky, Andrew Y. Ng. Learning syntactic patterns for automatic hypernym discovery • Sarkar. (1999). Applying Co-training methods to Statistical Parsing. • S. Brin, 1998. Extracting patterns and relations from the World Wide Web,  EDBT'98 • O. Etzioni et al., 2005. "Unsupervised Named-Entity Extraction from the Web: An Experimental Study," AI Journal, 2005.