Use of non-negative matrix factorization for language model adaptation in a lecture transcription task

Novák, Miroslav; Mammone, Richard J.

doi:10.1109/icassp.2001.940887

Cited by 19 publications

(16 citation statements)

References 3 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Several methods related to Latent Semantic Analysis (LSA) have been recently proposed as alternative paradigms for language model (LM) adaptation [1,6,10]. LSA, which was originally developed for information retrieval, tries to map an high-dimensional space, e.g.…”

Section: Introductionmentioning

confidence: 99%

“…In [7,6] a probabilistic LSA (PLSA) is instead proposed which maps document-word conditional distributions into topic mixtures with hidden variables. In [10] non-negative matrix factorization (NMF) [9] is applied, which has been recently proposed as an appealing alternative to SVD to represent semantic features of texts. This paper, proposed a LM adaptation scheme which combines the probabilistic LSA approach with the minimum discrimination information (MDI) estimation criterion.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Language model adaptation through topic decomposition and MDI estimation

Federico¹

2002

IEEE International Conference on Acoustics Speech and Signal Processing

View full text Add to dashboard Cite

This work presents a language model adaptation method combining the latent semantic analysis framework with the minimum discrimination information estimation criterion. In particular, an unsupervised topic model decomposition is built which allows to infer topic related word distributions from very short adaptation texts. The resulting word distribution is then used to constraint the estimation of a minimum divergence trigram language. With respect to previous work, implementation details are discussed that make such approach effective for a large scale application. Experimental results are provided for a digital library indexing task, i.e. the speech transcription of five historical documentary films. By adapting a trigram language model from very terse content descriptions, i.e. maximum ten words, available for each film, a word error rate relative reduction of 3.2% was achieved.

show abstract

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Language model adaptation through topic decomposition and MDI estimation

Federico¹

2002

IEEE International Conference on Acoustics Speech and Signal Processing

View full text Add to dashboard Cite

show abstract

“…In the paper [63], Novak and Mammone introduce a new method with NMF. In addition to the non-negativity, another property of this factorization is that the columns of W tend to represent groups of associated words.…”

Section: Speech Recognitionmentioning

confidence: 99%

Nonnegativity constraints in numerical analysis

Chen

Plemmons

2009

The Birth of Numerical Analysis

163

143

View full text Add to dashboard Cite

A survey of the development of algorithms for enforcing nonnegativity constraints in scientific computation is given. Special emphasis is placed on such constraints in least squares computations in numerical linear algebra and in nonlinear optimization. Techniques involving nonnegative low-rank matrix and tensor factorizations are also emphasized. Details are provided for some important classical and modern applications in science and engineering. For completeness, this report also includes an effort toward a literature survey of the various algorithms and applications of nonnegativity constraints in numerical analysis.

show abstract

“…NMF technique is an approach which lead to a part-based representation because only additive, not subtractive, combinations of the original data is allowed [1,2]. This technique has been used in many applications including classifying faces [5], dynamic positron emission tomography [6], image processing [7], and language model adaptation [11]. It was also applied to BSS for NNL model for famous bar problem [13] in reference [4], but failed to that for the extended bar problem presented in this paper in which the sources are more difficult to be separated.…”

Section: As =mentioning

confidence: 99%

Blind Source Separation with Pattern Expression NMF

Zhang

et al. 2006

Advances in Neural Networks - ISNN 2006

View full text Add to dashboard Cite

Abstract. Independent component analysis (ICA) is a widely applicable and effective approach in blind source separation (BSS) for basic ICA model, but with limitations that sources should be statistically independent, while more common situation is BSS for non-negative linear (NNL) model where observations are linear combinations of non-negative sources with non-negative coefficients and sources may be statistically dependent. By recognizing the fact that BSS for basic ICA model corresponds to matrix factorization problem, in this paper, a novel idea of BSS for NNL model is proposed that the BSS for NNL corresponds to a non-negative matrix factorization problem and the non-negative matrix factorization (NMF) technique is utilized. For better expression of the patterns of the sources, the NMF is further extended to pattern expression NMF (PE-NMF) and its algorithm is presented. Finally, the experimental results are presented which show the effectiveness and efficiency of the PE-NMF to BSS for a variety of applications which follow NNL model.

show abstract

Use of non-negative matrix factorization for language model adaptation in a lecture transcription task

Cited by 19 publications

References 3 publications

Language model adaptation through topic decomposition and MDI estimation

Language model adaptation through topic decomposition and MDI estimation

Nonnegativity constraints in numerical analysis

Blind Source Separation with Pattern Expression NMF

Contact Info

Product

Resources

About