Getting more from automatic transcripts for semi-supervised language modeling

Novotney, Scott; Schwartz, Richard; Khudanpur, Sanjeev

doi:10.1016/j.csl.2015.08.007

Cited by 5 publications

(4 citation statements)

References 23 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Latent Semantic Analysis (LSA) [7] and Latent Dirichlet Allocation (LDA) [8] are two widely used techniques for document retrieval, full-text retrieval [9], information retrieval [10] and topic modeling [11]. Similar techniques have been explored for topic identification and dynamic language model adaptation using vector space model [12], LSA [13], relevance language model [14], semi-supervised language models [15] and topic tracking language model [16]. LDA technique has been widely explored to form unsupervised adapted language model [17] and topic-specific language models for inflectional languages [18].…”

Section: Related Workmentioning

confidence: 99%

Efficient Search Mechanism from Large Scale Corpora for Domain-Specific Language Modeling in Speech Recognition

2019

ijeat

View full text Add to dashboard Cite

With the Internet and the World Wide Web revolution, large corpora in variety of forms are germinating ceaselessly that can be manifested as big data. One obligatory area for the usage of such large corpora is language modeling for large vocabulary continuous speech recognition. Language modeling is an indispensable module in speech recognition architecture, which plays a vital role in reducing the search space during the recognition process. Additionally, the language model that is contiguous to the domain of the speech can dwindle the search space and escalate the recognition accuracy. In this paper, an efficient searching mechanism for domain-specific document retrieval from the large corpora has been elucidated using Elasticsearch which is a distributed and an efficient search engine for big data. This assisted us in tuning the language model in accordance with the domain and also by reducing the search time by more than 90% in comparison to conventional search and retrieval mechanism used in our earlier work. A word level and a phrase level retrieval process for creating domain-specific language model has been implemented. The evaluation of the system is performed on the basis of word error rate (WER) and perplexity (PPL) of the speech recognition system. The results shows nearly 10% decrease on WER and a major reduction in the PPL that helped in boosting the performance of the speech recognition process. From the results, it can be consummated that Elasticsearch is an efficient mechanism for domain specific document retrieval from large corpora rather than using topic modeling toolkits

show abstract

Section: Related Workmentioning

confidence: 99%

Efficient Search Mechanism from Large Scale Corpora for Domain-Specific Language Modeling in Speech Recognition

2019

ijeat

View full text Add to dashboard Cite

show abstract

“…This shows a decrease of 1.5% in the WER on the MIT corpus and 1.3% decrease in WER on the corpus of spontaneous Japanese. Novotney et al [9] have explained the benefits of semi-supervised LMs for under-resourced languages and over a range of low resource condition. They have given limitation to back-off LMs and have motivated the robust use of automatic counts prior to the estimated parameters of a log-linear LM [9].…”

Section: Related Workmentioning

confidence: 99%

“…Novotney et al [9] have explained the benefits of semi-supervised LMs for under-resourced languages and over a range of low resource condition. They have given limitation to back-off LMs and have motivated the robust use of automatic counts prior to the estimated parameters of a log-linear LM [9]. Oger and Linarès [10] have used the possibility theory for adapting the LM.…”

Section: Related Workmentioning

confidence: 99%

See 1 more Smart Citation

Ameliorated language modelling for lecture speech recognition of Indian English

Phull

Kumar

2018

Sādhanā

View full text Add to dashboard Cite

A great amount of research is growing towards the automatic transcription of lectures that consist of numerous information and knowledge that could be helpful to the educational systems and institutes. In large vocabulary speech recognition, language model plays a paramount role in reducing the humongous search space. However, language modelling is very brittle when moving from one domain to another or when moving from read speech to spontaneous speech. Also, lecture speech recognition will have some of the characteristics of spontaneous speech. Hence, it is very challenging to build the language model for this task. In this paper, a judicious approach to adapt the language model in a way where the language model will be in close proximity to the topic spoken in the lecture speech has been depicted. The evaluation of the language model is devised using the proposed approach with the existing language models such as CMU Sphinx, Gigaword and HUB-4. We observed the results analysis that the language models devised from the proposed approach outperform from the existing language models in terms of word error rate, perplexity and out of vocabulary rate. Analysis shows that the presented two-phase approach has resulted in an average decrease of the word error rate to be approximately 14% and the perplexity is decreased by half on average.

show abstract