1 / 51

CS590I: Information Retrieval

CS590I: Information Retrieval . CS-590I Information Retrieval Retrieval Models: Language models Luo Si Department of Computer Science Purdue University. Retrieval Model: Language Model. Introduction to language model. Unigram language model. Document language model estimation.

neve-fry
Download Presentation

CS590I: Information Retrieval

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript

  1. CS590I: Information Retrieval • CS-590I • Information Retrieval • Retrieval Models: Language models • Luo Si • Department of Computer Science • Purdue University

  2. Retrieval Model: Language Model • Introduction to language model • Unigram language model • Document language model estimation • Maximum Likelihood estimation • Maximum a posterior estimation • Jelinek Mercer Smoothing • Model-based feedback

  3. Language Models: Motivation • Vector space model for information retrieval • Documents and queries are vectors in the term space • Relevance is measure by the similarity between document vectors and query vector • Problems for vector space model • Ad-hoc term weighting schemes • Ad-hoc similarity measurement No justification of relationship between relevance and similarity • We need more principled retrieval models…

  4. Introduction to Language Models: • Language model can be created for any language sample • A document • A collection of documents • Sentence, paragraph, chapter, query… • The size of language sample affects the quality of language model • Long documents have more accurate model • Short documents have less accurate model • Model for sentence, paragraph or query may not be reliable

  5. Introduction to Language Models: • A document language model defines a probability distribution over indexed terms • E.g., the probability of generating a term • Sum of the probabilities is 1 • A query can be seen as observed data from unknown models • Query also defines a language model (more on this later) • How might the models be used for IR? • Rank documents by Pr( | ) • Rank documents by language models of and based on kullback-Leibler (KL) divergence between the models (come later)

  6. sport, basketball Language Model for Language Model for Language Model for sport, basketball, ticket, sport stock, finance, finance, stock Language Model for IR: Example Generate retrieval results Estimate the generation probability of Pr( | ) Estimating language model for each document basketball, ticket, finance, ticket, sport

  7. Language Models • Three basic problems for language models • What type of probabilistic distribution can be used to construct language models? • How to estimate the parameters of the distribution of the language models? • How to compute the likelihood of generating queries given the language modes of documents?

  8. Examples: • Five words in vocabulary (sport, basketball, ticket, finance, stock) • For a document , its language mode is: • {Pi(“sport”), Pi(“basketball”), Pi(“ticket”), Pi(“finance”), Pi(“stock”)} Multinomial/Unigram Language Models • Language model built by multinomial distribution on single terms (i.e., unigram) in the vocabulary • Formally: • The language model is: {Pi(w) for any word w in vocabulary V}

  9. Multinomial Model for Multinomial Model for Multinomial Model for basketball, ticket, finance, ticket, sport sport, basketball, ticket, sport stock, finance, finance, stock Multinomial/Unigram Language Models Estimating language model for each document

  10. Maximum Likelihood Estimation (MLE) • Maximum Likelihood Estimation: • Find model parameters that make generation likelihood reach maximum: M*=argmaxMPr(D|M) There are K words in vocabulary, w1...wK (e.g., 5) Data: one document with counts tfi(w1), …, tfi(wK), and length | | Model: multinomial M with parameters {pi(wk)} Likelihood: Pr( | M) M*=argmaxMPr( |M)

  11. Maximum Likelihood Estimation (MLE) Use Lagrange multiplier approach Set partial derivatives to zero Get maximum likelihood estimate

  12. (psp, pb, pt, pf, pst) = (0.5,0.25,0.25,0,0) (psp, pb, pt, pf, pst) = (0.2,0.2,0.4,0.2,0) (psp, pb, pt, pf, pst) = (0,0,0,0.5,0.5) basketball, ticket, finance, ticket, sport sport, basketball, ticket, sport stock, finance, finance, stock Maximum Likelihood Estimation (MLE) Estimating language model for each document

  13. Maximum Likelihood Estimation (MLE) • Maximum Likelihood Estimation: • Assign zero probabilities to unseen words in small sample • A specific example: • Only two words in vocabulary (w1=sport, w2=business) like (head, tail) for a coin; A document generates sequence of two words or draw a coin for many times • Only observe two words (flip the coin twice) and MLE estimators are: • “business sport” Pi(w1)=0.5 • “sport sport” Pi(w1)=1 ? • “business business” Pi(w1)=0 ?

  14. Maximum Likelihood Estimation (MLE) • A specific example: • Only observe two words (flip the coin twice) and MLE estimators are: • “business sport” Pi(w1)*=0.5 • “sport sport” Pi(w1)*=1 ? • “business business” Pi(w1)*=0 ? • Data sparseness problem

  15. Solution to Sparse Data Problems • Maximum a posterior (MAP) estimation • Shrinkage • Bayesian ensemble approach

  16. A specific examples: • Only two words in vocabulary (sport, business) • For a document : Prior Distribution Maximum A Posterior (MAP) Estimation • Maximum A Posterior Estimation: • Select a model that maximizes the probability of model given observed data M*=argmaxMPr(M|D)=argmaxMPr(D|M)Pr(M) • Pr(M): Prior belief/knowledge • Use prior Pr(M) to avoid zero probabilities

  17. Maximum A Posterior (MAP) Estimation • Maximum A Posterior Estimation: • Introduce prior on the multinomial distribution • Use prior Pr(M) to avoid zero probabilities, most of coins are more or less unbiased • Use Dirichlet prior on p(w) Hyper-parameters Constant for pK (x) is gamma function

  18. For the two word example: • a Dirichlet prior Maximum A Posterior (MAP) Estimation P(w1)2 (1-P(w1)2)

  19. Maximum A Posterior (MAP) Estimation • Maximum A Posterior: M*=argmaxMPr(M|D)=argmaxMPr(D|M)Pr(M) Pseudo Counts

  20. Maximum A Posterior (MAP) Estimation • A specific example: • Only observe two words (flip a coin twice): • “sport sport” Pi(w1)*=1 ? P(w1)2 (1-P(w1)2) times

  21. Maximum A Posterior (MAP) Estimation • A specific example: • Only observe two words (flip a coin twice): • “sport sport” Pi(w1)*=1 ?

  22. MAP EstimationUnigram Language Model • Maximum A Posterior Estimation: • Use Dirichlet prior for multinomial distribution • How to set the parameters for Dirichlet prior

  23. MAP EstimationUnigram Language Model • Maximum A Posterior Estimation: • Use Dirichlet prior for multinomial distribution • There are K terms in the vocabulary: Hyper-parameters Constant for pK

  24. MAP EstimationUnigram Language Model • MAP Estimation for unigram language model: • Use Lagrange Multiplier; Set derivative to 0 Pseudo counts set by hyper-parameters

  25. MAP EstimationUnigram Language Model • MAP Estimation for unigram language model: • Use Lagrange Multiplier; Set derivative to 0 • How to determine the appropriate value for hyper-parameters? • When nothing observed from a document • What is most likely pi(wk) without looking at the content of the document?

  26. MAP EstimationUnigram Language Model • MAP Estimation for unigram language model: • What is most likely pi(wk) without looking at the content of the document? • The most likely pi(wk) without looking into the content of the document d is the unigram probability of the collection: • {p(w1|c), p(w2|c),…, p(wK|c)} • Without any information, guess the behavior of one member on the behavior of whole population Constant

  27. MAP EstimationUnigram Language Model • MAP Estimation for unigram language model: • Use Lagrange Multiplier; Set derivative to 0 Pseudo counts Pseudo document length

  28. Maximum A Posterior (MAP) Estimation • Dirichlet MAP Estimation for unigram language model: • Step 0: compute the probability on whole collection based collection unigram language model • Step 1: for each document , compute its smoothed unigram language model (Dirichlet smoothing) as

  29. Maximum A Posterior (MAP) Estimation • Dirichlet MAP Estimation for unigram language model: • Step 2: For a given query ={tfq(w1),…, tfq(wk)} • For each document , compute likelihood • The larger the likelihood, the more relevant the document is to the query

  30. Dirichlet Smoothing & TF-IDF • Dirichlet Smoothing: ? • TF-IDF Weighting:

  31. Dirichlet Smoothing & TF-IDF • Dirichlet Smoothing: • TF-IDF Weighting:

  32. Dirichlet Smoothing & TF-IDF • Dirichlet Smoothing:

  33. Dirichlet Smoothing & TF-IDF • Dirichlet Smoothing: Irrelevant part • TF-IDF Weighting:

  34. Dirichlet Smoothing & TF-IDF • Dirichlet Smoothing: • Look at the tf.idf part

  35. Dirichlet Smoothing Hyper-Parameter • Dirichlet Smoothing: Hyper-parameter • When is very small, approach MLE estimator • When is very large, approach probability on whole collection • How to set appropriate ?

  36. w1 Leave w1 out Leave wj out ... wj ... Dirichlet Smoothing Hyper-Parameter • Leave One out Validation: ... ...

  37. w1 ... wj ... Dirichlet Smoothing Hyper-Parameter • Leave One out Validation: Leave all words out one by one for a document Do the procedure for all documents in a collection Find appropriate

  38. Dirichlet Smoothing Hyper-Parameter • What type of document/collection would get large ? • Most documents use similar vocabulary and wording pattern as the whole collection • What type of document/collection would get small ? • Most documents use different vocabulary and wording pattern than the whole collection

  39. U.S. Indiana West Lafayette Shrinkage • Maximum Likelihood (MLE) builds model purely on document data and generates query word • Model may not be accurate when document is short (many unseen words) • Shrinkage estimator builds more reliable model by consulting more general models (e.g., collection language model) Example: Estimate P(Lung_Cancer|Smoke)

  40. Jelinek Mercer Smoothing • Assume for each word, with probability , it is generated from document language model (MLE), with probability 1- , it is generated from collection language model (MLE) • Linear interpolation between document language model and collection language model Shrinkage JM Smoothing:

  41. Shrinkage • Relationship between JM Smoothing and Dirichlet Smoothing JM Smoothing:

  42. Model Based Feedback • Equivalence of retrieval based on query generation likelihood and Kullback-Leibler (KL) Divergence between query and document language models • Kullback-Leibler (KL) Divergence between two probabilistic distributions • It is the distance between two probabilistic distributions • It is always larger than zero How to prove it ?

  43. Model Based Feedback • Equivalence of retrieval based on query generation likelihood and Kullback-Leibler (KL) Divergence between query and document language models Loglikelihood of query generation probability Document independent constant • Generalize query representation to be a distribution (fractional term weighting)

  44. Retrieval results Retrieval results Calculate KL Divergence Estimate the generation probability of Pr( | ) Language Model for Language Model for Language Model for Estimating query language model Estimating language model Estimating document language model Model Based Feedback

  45. Language Model for Language Model for Language Model for No feedback Full feedback Estimating document language model Model Based Feedback Feedback Documents from initial results Retrieval results Calculate KL Divergence New Query Model Estimating query language model

  46. Model Based Feedback: Estimate • Assume there is a generative model to produce each word within feedback document(s) • For each word in feedback document(s), given  Background model Feedback Documents 1- PC(w) w Flip a coin Topic words  w qF(w)

  47. Model Based Feedback: Estimate MLE Estimator • For each word, there is a hidden variable telling which language model it comes from the 0.12 to 0.05 it 0.04 a 0.02 … sport 0.0001 basketball 0.00005 … Feedback Documents Background Model pC(w|C) 1-=0.8 Unknown query topic p(w|F)=? “Basketball” … sport =? basketball =? game =? player =? … =0.2 If we know the value of hidden variable of each word ...

  48. Model Based Feedback: Estimate E-step • For each word, the hidden variable Zi = {1 (feedback), 0 (background} • Step1: estimate hidden variable based current on model parameter (Expectation) the (0.1) basketball (0.7) game (0.6) is (0.2) …. • Step2: Update model parameters based on the guess in step1 (Maximization) M-Step

  49. Model Based Feedback: Estimate • Expectation-Maximization (EM) algorithm • Step 0: Initialize values of • Step1: (Expectation) • Step2: (Maximization) Give =0.5

  50. Model Based Feedback: Estimate • Properties of parameter  • If is close to 0, most common words can be generated from collection language model, so more topic words in query language mode • If is close to 1, query language model has to generate most common words, so fewer topic words in query language mode

More Related