Topic extraction based on continuous speech recognition in broadcast news speech

IEICE Transactions on Information and Systems

Volumn E85-D, Issue 7, 2002, Pages 1138-1144

  Ohtsuki, Katsutoshi ^a   Matsuoka, Tatsuo ^b   Matsunaga, Shoichi ^a   Furui, Sadaoki ^c  

^a NTT CORPORATION   (Japan)

^b Sapporo Medical Center   (Japan)

^c TOKYO INSTITUTE OF TECHNOLOGY   (Japan)

Author keywords

Broadcast news; Continuous speech recognition; Relevance score; Topic extraction; Topic word

Indexed keywords

BROADCASTING; DATABASE SYSTEMS; MAPPING; MATHEMATICAL MODELS; STATISTICAL METHODS; WORD PROCESSING;

TOPIC EXTRACTION;

CONTINUOUS SPEECH RECOGNITION;

EID: 0036630943     PISSN: 09168532     EISSN: None     Source Type: Journal    
DOI: None     Document Type: Article

References (15)

1. Talker independent keyword spotting for information retrieval
2. A system for unrestricted topic retrieval form radio news broadcasts
3. Issues in topic identification on the switchboard corpus
4. Techniques for information retrieval from voice messages
5. Spotting events in continuous speech
6. Improvements in switchboard recognition and topic identification
7. Gisting conversational speech
8. Improved topic discrimination of broadcast news using a model of multiple simultaneous topics
9. Topic extraction based on continuous speech recognition in broadcast-news speech
10. Topic extraction with multiple topic-words in broadcast-news speech
11. Toward automatic transcription of Japanese broadcast news
12. Toward automatic transcription of Japanese broadcast news
13. Japanese large-vocabulary continuous-speech recognition using a newspaper corpus and broadcast news
14. Tree-based state tying for high accuracy acoustic modeling
15. Estimation of probabilities from sparse data for the language model component of a speech recognizer