On the estimation of 'small' probabilities by leaving-one-out

Ney, Hermann; Essen, Ute; Kneser, Reinhard

doi:10.1109/34.476512

Cited by 66 publications

(64 citation statements)

References 13 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Our framework then enriches the user language model with social network information. We select four popular smoothing methods to demonstrate such effect, namely additive smoothing, absolute smoothing (Ney et al, 1995), Jelinek-Mercer smoothing (Jelinek and Mercer, 1980) and Dirichlet smoothing (MacKay and Peto, 1994). The results of using only the base model (i.e.…”

Section: Baseline Methodsmentioning

confidence: 99%

Enriching Cold Start Personalized Language Model Using Social Network Information

Huang

Yan

Kuo

et al. 2014

Proceedings of the 52nd Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers)

View full text Add to dashboard Cite

We introduce a generalized framework to enrich the personalized language models for cold start users. The cold start problem is solved with content written by friends on social network services. Our framework consists of a mixture language model, whose mixture weights are estimated with a factor graph. The factor graph is used to incorporate prior knowledge and heuristics to identify the most appropriate weights. The intrinsic and extrinsic experiments show significant improvement on cold start users.

show abstract

Section: Baseline Methodsmentioning

confidence: 99%

Enriching Cold Start Personalized Language Model Using Social Network Information

Huang

Yan

Kuo

et al. 2014

Proceedings of the 52nd Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers)

View full text Add to dashboard Cite

show abstract

“…For this purpose, one should note that, in this kind of deterministic models, the likelihood of the training sample is maximized if the stochastic model assigns to every tree t in the sample a probability equal to its relative frequency in Ω [8]. So, these probabilities must be calculated as the ratio of the number of occurrences of a transition to the number of occurrences of the state to which this transition leads.…”

Section: Stochastic K-testable Tree Modelsmentioning

confidence: 99%

“…Therefore, the use of smoothing techniques becomes necessary if one wants to use these models for parsing. Two classical techniques of this type are linear interpolation and backing-off [8].…”

Section: Smoothingmentioning

confidence: 99%

Classifying Melodies Using Tree Grammars

Bernabeu

Calera-Rubio

Iñesta

2011

Pattern Recognition and Image Analysis

View full text Add to dashboard Cite

Abstract. Similarity computation is a difficult issue in music information retrieval, because it tries to emulate the special ability that humans show for pattern recognition in general, and particularly in the presence of noisy data. A number of works have addressed the problem of what is the best representation for symbolic music in this context. The tree representation, using rhythm for defining the tree structure and pitch information for leaf and node labeling has proven to be effective in melodic similarity computation. In this paper we propose a solution when we have melodies represented by trees for the training but the duration information is not available for the input data. For that, we infer a probabilistic context-free grammar using the information in the trees (duration and pitch) and classify new melodies represented by strings using only the pitch. The case study in this paper is to identify a snippet query among a set of songs stored in symbolic format. For it, the utilized method must be able to deal with inexact queries and efficient for scalability issues.

show abstract

“…We include the plain smoothing of Additive (also known as Add-δ) smoothing and Absolute Discounting decrease the probability of seen words by subtracting a constant (Ney et al, 1995). We also implement several classic strategies smoothed from the whole collection as background information: Jelinek-Mercer (J-M) applies a linear interpolation, and Dirichlet employs a prior on collection influence .…”

Section: Algorithms For Comparisonmentioning

confidence: 99%

Tackling Sparsity, the Achilles Heel of Social Networks: Language Model Smoothing via Social Regularization

Yan

Liu

et al. 2015

Proceedings of the 53rd Annual Meeting of the Association for Computational Linguistics and the 7th International Joint Confere

View full text Add to dashboard Cite

Online social networks nowadays have the worldwide prosperity, as they have revolutionized the way for people to discover, to share, and to diffuse information. Social networks are powerful, yet they still have Achilles Heel: extreme data sparsity. Individual posting documents, (e.g., a microblog less than 140 characters), seem to be too sparse to make a difference under various scenarios, while in fact they are quite different. We propose to tackle this specific weakness of social networks by smoothing the posting document language model based on social regularization. We formulate an optimization framework with a social regularizer. Experimental results on the Twitter dataset validate the effectiveness and efficiency of our proposed model.

show abstract

On the estimation of 'small' probabilities by leaving-one-out

Cited by 66 publications

References 13 publications

Enriching Cold Start Personalized Language Model Using Social Network Information

Enriching Cold Start Personalized Language Model Using Social Network Information

Classifying Melodies Using Tree Grammars

Tackling Sparsity, the Achilles Heel of Social Networks: Language Model Smoothing via Social Regularization

Contact Info

Product

Resources

About