PATTERN RECOGNITION in SPEECH and LANGUAGE PROCESSING Edited by WU CHOU Avaya Labs Research BIING HWANG JUANG Georgia Institute of Technology CRC PR E S S Boca Raton London New York Washington, D.C Preface Basking Ridge, New Jersey September, 2002 Contributors A Abella Vaibhava Goel James Allan Allen L Gorin T Alonso Qiang Huo Jerome R Bellegarda Biing-Hwang Juang William Byrne Shigeru Katagiri Wu Chou Sadaoki Furui Lori Lamel Jean-Luc Gauvain Qi (Peter) Li G Riccardi John Makhoul Richard M Schwartz Hermann Ney J H Wright F J Och Contents Minimum Classification Error (MCE) Approach in Pattern Recognition Wu Chou Minimum Bayes-Risk Methods in Automatic Speech Recognition Ý Vaibhava Goel £ and William ByrneÝ £ A Decision Theoretic Formulation for Robust Automatic Speech Recognition Qiang Huo Speech Pattern Recognition using Neural Networks Shigeru Katagiri Large Vocabulary Speech Recognition Based on Statistical Methods Jean-Luc Gauvain and Lori Lamel Toward Spontaneous Speech Recognition and Understanding Sadaoki Furui 0.02 90 0.10.2 0.5 10 20 40 60 80 80 60 60 40 40 20 20 10 10 Miss Rate 80 2 0.02 0.10.2 0.5 10 20 False Alarm Rate 40 60 80 90 FIGURE 12.1 A sample detection error tradeoff (DET) curve for the TDT tracking task with one training story ( ỉ ẵ) ặ minimum 12.2 Basic Topic Models 12.2.1 Vector Space Ù¡Ú Ù ¾ Ă ắ 12.2.2 Language Models ẩ Ê Ê ẩ ẩ ẵ ắ à ắ ắ ẵ 12.3 Implementing the Models È ´Û Å µ ŵ 12.3.1 Named Entities President Bush George Bush 12.3.2 Document Expansion È ´Û Å µ 12.3.3 Clustering Å ¾Å È ´Å µÈ ´Û Å µ 12.3.4 Time Decay 12.4 Comparing Models 12.4.1 Nearest Neighbors ẵ 12.4.2 Decision Trees 12.4.3 Model-to-Model ị ễ ếà ễ ế ễ ếà ĩắ ĩà ễĩà éể ễĩà ế ắà à ắ ẵà ẵ ắà ẵ ĩ 12.5 Miscellaneous Issues ị ĩ ẵ ắ ẵ ẵ 12.5.1 Deferral 12.5.2 Multi-modal Issues third 12.5.3 Multi-lingual Issues FIGURE 12.2 Screen snapshot of the Lighthouse system that was created to portray TDT topic clusters and their relationships 12.6 Using TDT Interactively 12.6.1 Demonstrations 12.6.2 Timelines ¾ Oklahoma ¾ Oklahoma McVeigh Simpson FIGURE 12.3 Overview of January-June 1998 The topic labeled monica lewinsky allegation is the highest ranked topic by the ¾ measure The pop-up on oregon school shooting shows significant named entities for that event The other pop-up displays a sub-menu for obtaining more information on the name kip kinkel ¾ 12.7 Modeling Events ß 12.8 Conclusion ß research References Proceedings of Conference on Information Retrieval Research (SIGIR) Proceedings of the DARPA Broadcast News Transcription and Understanding Workshop Proceedings of Conference on Information Retrieval Research (SIGIR) Information Retrieval Topic Detection and Tracking: Event-based Information Organization In Proceedings of the 36th Annual Meeting of the Association for Computational Linguistics and the 17th International Conference on Computational Linguistics (COLING-ACL’98) Proceedings for Empirical Methods in NLP .. .PATTERN RECOGNITION in SPEECH and LANGUAGE PROCESSING Edited by WU CHOU Avaya Labs Research BIING HWANG JUANG Georgia Institute of Technology CRC PR E S S Boca Raton London New York Washington,... using Neural Networks Shigeru Katagiri Large Vocabulary Speech Recognition Based on Statistical Methods Jean-Luc Gauvain and Lori Lamel Toward Spontaneous Speech Recognition and Understanding... Bayes-Risk Methods in Automatic Speech Recognition Ý Vaibhava Goel £ and William ByrneÝ £ A Decision Theoretic Formulation for Robust Automatic Speech Recognition Qiang Huo Speech Pattern Recognition