자동 구두점 삽입을 이용한 Rich Transcription 생성 = Rich Transcription Generation Using Automatic Insertion of Punctuation Marks

다국어 초록 (Multilingual Abstract)

A punctuation generation system which combines prosodic information with acoustic and language model information is presented. Experiments have been conducted first for the reference text transcriptions. In these experiments, prosodic information was shown to be more useful than language model information. When these information sources are combined, an F-measure of up to 0.7830 was obtained for adding punctuation to a reference transcription. This method of punctuation generation can also be applied to the 1-best output of a speech recogniser. The 1-best output is first time aligned. Based on the time alignment information, prosodic features are generated. As in the approach applied in the punctuation generation for reference transcriptions, the best sequence of punctuation marks for this 1-best output is found using the prosodic feature model and an language model trained on texts which contain punctuation marks.

참고문헌 (Reference)

1 P. Woodland, "The development of the HTK broadcast news transcription system: An overview" 37 (37): 47-67, 2002

2 J. Odell, "The CUHTK-Entropic 10xRT broadcast news transcription system" 271-275, 1999

3 P. Woodland, "The 1998 HTK broadcast news transcription system: Development and results" 265-270, 1999

4 P. Woodland, "The 1997 HTK broadcast news transcription system" 1998

5 C. Chen, "Speech recognition with automatic punctuation" 447-450, 1999

6 Y. Gotoh, "Sentence boundary detection in broadcast speech transcripts" 228-235, 2000

7 J. Makhoul, "Performance measures for information extraction" 249-252, 1999

8 C. Leggetter, "Maximum likelihood linear regression for speaker adaptation of continuous density hidden Markov models" 9 : 171-185, 1995

9 P. Taylor, "Intonation and dialog context as constraints for speech recognition" 41 (41): 489-508, 1998

10 "Fundamentals of speech recognition" Prentice Hall 1993

11 D. Beeferman, "Cyberpunc: A lightweight punctuation annotation system for speech" 689-692, 1998

12 T. Niesler, "Comparison of part-of-speech and automatically derived category-based language models for speech recognition" 177-180, 1998

13 A. Stolcke, "Combining words and speech prosody for automatic topic segmentation" 61-64, 1999

14 D. Hakkani-Tur, "Combining words and prosody for information extraction from speech" 1991-1994, 1999

15 S. Abney, "Chunks and dependencies: Bringing processing evidence to bear on syntax" 145-164, 1995

16 E. Shriberg, "Can prosody aid the automatic classification of dialog acts in conversational speech?" 41 (41): 439-487, 1998

17 D. Palmer, "Adaptive multilingual sentence boundary disambiguation" 23 (23): 241-269, 1997

18 Silverman, "A standard for labelling English prosody”" 867-870, 1992

19 M. Fach, "A comparison between syntactic and prosodic phrasing" 527-530, 1999

20 D. Pallett, "1998 broadcast news benchmark test results: English and non-English word error rate performance measures" 5-12, 1999