Results 1  10
of
93
An Empirical Study of Smoothing Techniques for Language Modeling
, 1998
"... We present an extensive empirical comparison of several smoothing techniques in the domain of language modeling, including those described by Jelinek and Mercer (1980), Katz (1987), and Church and Gale (1991). We investigate for the first time how factors such as training data size, corpus (e.g., Br ..."
Abstract

Cited by 1188 (21 self)
 Add to MetaCart
(Show Context)
We present an extensive empirical comparison of several smoothing techniques in the domain of language modeling, including those described by Jelinek and Mercer (1980), Katz (1987), and Church and Gale (1991). We investigate for the first time how factors such as training data size, corpus (e.g., Brown versus Wall Street Journal), and ngram order (bigram versus trigram) affect the relative performance of these methods, which we measure through the crossentropy of test data. In addition, we introduce two novel smoothing techniques, one a variation of JelinekMercer smoothing and one a very simple linear interpolation technique, both of which outperform existing methods. 1
The Infinite Hidden Markov Model
 Machine Learning
, 2002
"... We show that it is possible to extend hidden Markov models to have a countably infinite number of hidden states. By using the theory of Dirichlet processes we can implicitly integrate out the infinitely many transition parameters, leaving only three hyperparameters which can be learned from data. Th ..."
Abstract

Cited by 631 (41 self)
 Add to MetaCart
We show that it is possible to extend hidden Markov models to have a countably infinite number of hidden states. By using the theory of Dirichlet processes we can implicitly integrate out the infinitely many transition parameters, leaving only three hyperparameters which can be learned from data. These three hyperparameters define a hierarchical Dirichlet process capable of capturing a rich set of transition dynamics. The three hyperparameters control the time scale of the dynamics, the sparsity of the underlying statetransition matrix, and the expected number of distinct hidden states in a finite sequence. In this framework it is also natural to allow the alphabet of emitted symbols to be infiniteconsider, for example, symbols being possible words appearing in English text.
A Gaussian prior for smoothing maximum entropy models
, 1999
"... In certain contexts, maximum entropy (ME) modeling can be viewed as maximum likelihood training for exponential models, and like other maximum likelihood methods is prone to overfitting of training data. Several smoothing methods for maximum entropy models have been proposed to address this problem ..."
Abstract

Cited by 252 (2 self)
 Add to MetaCart
(Show Context)
In certain contexts, maximum entropy (ME) modeling can be viewed as maximum likelihood training for exponential models, and like other maximum likelihood methods is prone to overfitting of training data. Several smoothing methods for maximum entropy models have been proposed to address this problem, but previous results do not make it clear how these smoothing methods compare with smoothing methods for other types of related models. In this work, we survey previous work in maximum entropy smoothing and compare the performance of several of these algorithms with conventional techniques for smoothing ngram language models. Because of the mature body of research in ngram model smoothing and the close connection between maximum entropy and conventional ngram models, this domain is wellsuited to gauge the performance of maximum entropy smoothing methods. Over a large number of data sets, we find that an ME smoothing method proposed to us by Lafferty [1] performs as well as or better than all other algorithms under consideration. This general and efficient method involves using a Gaussian prior on the parameters of the model and selecting maximum a posteriori instead of maximum likelihood parameter values. We contrast this method with previous ngram smoothing methods to explain its superior performance.
A fully bayesian approach to unsupervised partofspeech tagging
 In ACL
, 2007
"... Unsupervised learning of linguistic structure is a difficult problem. A common approach is to define a generative model and maximize the probability of the hidden structure given the observed data. Typically, this is done using maximumlikelihood estimation (MLE) of the model parameters. We show usi ..."
Abstract

Cited by 164 (2 self)
 Add to MetaCart
Unsupervised learning of linguistic structure is a difficult problem. A common approach is to define a generative model and maximize the probability of the hidden structure given the observed data. Typically, this is done using maximumlikelihood estimation (MLE) of the model parameters. We show using partofspeech tagging that a fully Bayesian approach can greatly improve performance. Rather than estimating a single set of parameters, the Bayesian approach integrates over all possible parameter values. This difference ensures that the learned structure will have high probability over a range of possible parameters, and permits the use of priors favoring the sparse distributions that are typical of natural language. Our model has the structure of a standard trigram HMM, yet its accuracy is closer to that of a stateoftheart discriminative model (Smith and Eisner, 2005), up to 14 percentage points better than MLE. We find improvements both when training from data alone, and using a tagging dictionary. 1
Topic modeling: beyond bagofwords
 NIPS 2005 Workshop on Bayesian Methods for Natural Language Processing
, 2005
"... Some models of textual corpora employ text generation methods involving ngram statistics, while others use latent topic variables inferred using the “bagofwords ” assumption, in which word order is ignored. Previously, these methods have not been combined. In this work, I explore a hierarchical g ..."
Abstract

Cited by 151 (4 self)
 Add to MetaCart
(Show Context)
Some models of textual corpora employ text generation methods involving ngram statistics, while others use latent topic variables inferred using the “bagofwords ” assumption, in which word order is ignored. Previously, these methods have not been combined. In this work, I explore a hierarchical generative probabilistic model that incorporates both ngram statistics and latent topic variables by extending a unigram topic model to include properties of a hierarchical Dirichlet bigram language model. The model hyperparameters are inferred using a Gibbs EM algorithm. On two data sets, each of 150 documents, the new model exhibits better predictive accuracy than either a hierarchical Dirichlet bigram language model or a unigram topic model. Additionally, the inferred topics are less dominated by function words than are topics discovered using unigram statistics, potentially making them more meaningful. 1.
Maximum Likelihood and Covariant Algorithms for Independent Component Analysis
, 1996
"... Bell and Sejnowski (1995) have derived a blind signal processing algorithm for a nonlinear feedforward network from an information maximization viewpoint. This paper first shows that the same algorithm can be viewed as a maximum likelihood algorithm for the optimization of a linear generative model ..."
Abstract

Cited by 117 (1 self)
 Add to MetaCart
Bell and Sejnowski (1995) have derived a blind signal processing algorithm for a nonlinear feedforward network from an information maximization viewpoint. This paper first shows that the same algorithm can be viewed as a maximum likelihood algorithm for the optimization of a linear generative model. Second, a covariant version of the algorithm is derived. This algorithm is simpler and somewhat more biologically plausible, involving no matrix inversions; and it converges in a smaller number of iterations. Third, this paper gives a partial proof of the `folktheorem' that any mixture of sources with highkurtosis histograms is separable by the classic ICA algorithm. Fourth, a collection of formulae are given that may be useful for the adaptation of the nonlinearity in the ICA algorithm. 1 Blind separation Algorithms for blind separation (Jutten and Herault 1991; Comon et al. 1991; Bell and Sejnowski 1995; Hendin et al. 1994) attempt to recover source signals s from observations x whic...
A Bit of Progress in Language Modeling
, 2001
"... Language modeling is the art of determining the probability of a sequence of words. This is useful in a large variety of areas including speech recognition, optical character recognition, handwriting recognition, machine translation, and spelling correction (Church, 1988; Brown et al., 1990; Hull, 1 ..."
Abstract

Cited by 114 (2 self)
 Add to MetaCart
Language modeling is the art of determining the probability of a sequence of words. This is useful in a large variety of areas including speech recognition, optical character recognition, handwriting recognition, machine translation, and spelling correction (Church, 1988; Brown et al., 1990; Hull, 1992; Kernighan et al., 1990; Srihari and Baltus, 1992). The most commonly used language models are very simple (e.g. a Katzsmoothed trigram model). There are many improvements over this simple model however, including caching, clustering, higherorder ngrams, skipping models, and sentencemixture models, all of which we will describe below. Unfortunately, these more complicated techniques have rarely been examined in combination. It is entirely possible that two techniques that work well separately will not work well together, and, as we will show, even possible that some techniques will work better together than either one does by itself. In this...
A bayesian framework for word segmentation: Exploring the effects of context
 In 46th Annual Meeting of the ACL
, 2009
"... Since the experiments of Saffran et al. (1996a), there has been a great deal of interest in the question of how statistical regularities in the speech stream might be used by infants to begin to identify individual words. In this work, we use computational modeling to explore the effects of differen ..."
Abstract

Cited by 110 (30 self)
 Add to MetaCart
Since the experiments of Saffran et al. (1996a), there has been a great deal of interest in the question of how statistical regularities in the speech stream might be used by infants to begin to identify individual words. In this work, we use computational modeling to explore the effects of different assumptions the learner might make regarding the nature of words – in particular, how these assumptions affect the kinds of words that are segmented from a corpus of transcribed childdirected speech. We develop several models within a Bayesian ideal observer framework, and use them to examine the consequences of assuming either that words are independent units, or units that help to predict other units. We show through empirical and theoretical results that the assumption of independence causes the learner to undersegment the corpus, with many two and threeword sequences (e.g. what’s that, do you, in the house) misidentified as individual words. In contrast, when the learner assumes that words are predictive, the resulting segmentation is far more accurate. These results indicate that taking context into account is important for a statistical word segmentation strategy to be successful, and raise the possibility that even young infants may be able to exploit more subtle statistical patterns than have usually been considered. 1
A survey of smoothing techniques for ME models
 IEEE Transactions on Speech and Audio Processing
, 2000
"... Abstract—In certain contexts, maximum entropy (ME) modeling can be viewed as maximum likelihood (ML) training for exponential models, and like other ML methods is prone to overfitting of training data. Several smoothing methods for ME models have been proposed to address this problem, but previous ..."
Abstract

Cited by 104 (1 self)
 Add to MetaCart
(Show Context)
Abstract—In certain contexts, maximum entropy (ME) modeling can be viewed as maximum likelihood (ML) training for exponential models, and like other ML methods is prone to overfitting of training data. Several smoothing methods for ME models have been proposed to address this problem, but previous results do not make it clear how these smoothing methods compare with smoothing methods for other types of related models. In this work, we survey previous work in ME smoothing and compare the performance of several of these algorithms with conventional techniques for smoothinggram language models. Because of the mature body of research ingram model smoothing and the close connection between ME and conventionalgram models, this domain is wellsuited to gauge the performance of ME smoothing methods. Over a large number of data sets, we find that fuzzy ME smoothing performs as well as or better than all other algorithms under consideration. We contrast this method with previousgram smoothing methods to explain its superior performance. Index Terms—Exponential models, language modeling, maximum entropy, minimum divergence,gram models, smoothing.