Automatic query generation and query relevance measurement for unsupervised language model adaptation of speech recognition

Authors:
Akinori Ito;Yasutomo Kajiura;Motoyuki Suzuki;Shozo Makino
Affiliations:
Graduate School of Engineering, Tohoku University, Sendai, Japan;Graduate School of Engineering, Tohoku University, Sendai, Japan;Institute of Technology and Science, University of Tokushima, Tokushima, Tokushima, Japan;Graduate School of Engineering, Tohoku University, Sendai, Japan
Venue:
EURASIP Journal on Audio, Speech, and Music Processing
Year:
2009

Citing 6
Cited 0

Document Categorization and Query Generation on the World Wide WebUsing WebACE

Artificial Intelligence Review - Special issue on data mining on the Internet
Unsupervised learning by probabilistic latent semantic analysis

Machine Learning
Information Retrieval

Information Retrieval
Enhanced word clustering for hierarchical text classification

Proceedings of the eighth ACM SIGKDD international conference on Knowledge discovery and data mining
Word sense disambiguation in information retrieval revisited

Proceedings of the 26th annual international ACM SIGIR conference on Research and development in informaion retrieval
Language model adaptation for fixed phrases by amplifying partial n-gram sequences

Systems and Computers in Japan

Quantified Score

Hi-index	0.00

Visualization

Abstract

We are developing a method of Web-based unsupervised language model adaptation for recognition of spoken documents. The proposed method chooses keywords from the preliminary recognition result and retrieves Web documents using the chosen keywords. A problem is that the selected keywords tend to contain misrecognized words. The proposed method introduces two new ideas for avoiding the effects of keywords derived from misrecognized words. The first idea is to compose multiple queries from selected keyword candidates so that the misrecognized words and correct words do not fall into one query. The second idea is that the number of Web documents downloaded for each query is determined according to the "query relevance." Combining these two ideas, we can alleviate bad effect of misrecognized keywords by decreasing the number of downloaded Web documents from queries that contain misrecognized keywords. Finally, we examine a method of determining the number of iterative adaptations based on the recognition likelihood. Experiments have shown that the proposed stopping criterion can determine almost the optimum number of iterations. In the final experiment, the word accuracy without adaptation (55.29%) was improved to 60.38%, which was 1.13 point better than the result of the conventional unsupervised adaptation method (59.25%).