| Presentation | 1996/3/21 On Some Linguistic and Prosodic Elements in the Final Part of the Preceeding Utterance that are Relevant to the Speakerhood of the Next Utterance Fragment Hanae Koiso, Yasuo Horiuchi, Syun Tutiya, Akira Ichikawa, |
|---|---|
| PDF Download Page | PDF download Page Link |
| Abstract(in Japanese) | (See Japanese page) |
| Abstract(in English) | It has been suggested in the literature that there are some determinant elements which are either grammatical, lexical, prosodic or paralinguistic that signal which the speaker of an utterance fragment intends to be the speaker of the next utterance fragment, herself or the interlocuter. In this study, we singled out a set of relevant 3 factors that are jointly determinant of the intended speakership or the next utterance fragment. They are lexical (P0S) and prosodic (pitch and power level). Certain lexical types such as sentence-final particles and intensifiers, are effectively determinant of who speaks next. Others are not. But prosody and power are considered to significantly contribute to the determination. We have been able to model the method the speaker signals how she controls her discourse, viz. to terminate or to continue. The proposed model is a weighted and normalized sum of the 3 factors as analyzed above. The model fits the corpus of 3 minutes map task dialogue in Japanese at the rate of 96.2% in the case of continuation and 92.9% in the case of termination. The rest can be explained to some extent but also suggest the limitations of the model. |
| Keyword(in Japanese) | (See Japanese page) |
| Keyword(in English) | speaker shift / Chiba Map Task Dialogue Corpus / task-oriented dialogue / spontaneous spoken dialogue / prosodic features |
| Paper # | NLC95-72 |
| Date of Issue |
| Conference Information | |
| Committee | NLC |
|---|---|
| Conference Date | 1996/3/21(1days) |
| Place (in Japanese) | (See Japanese page) |
| Place (in English) | |
| Topics (in Japanese) | (See Japanese page) |
| Topics (in English) | |
| Chair | |
| Vice Chair | |
| Secretary | |
| Assistant | |
| Paper Information | |
| Registration To | Natural Language Understanding and Models of Communication (NLC) |
|---|---|
| Language | JPN |
| Title (in Japanese) | (See Japanese page) |
| Sub Title (in Japanese) | (See Japanese page) |
| Title (in English) | On Some Linguistic and Prosodic Elements in the Final Part of the Preceeding Utterance that are Relevant to the Speakerhood of the Next Utterance Fragment |
| Sub Title (in English) | |
| Keyword(1) | speaker shift |
| Keyword(2) | Chiba Map Task Dialogue Corpus |
| Keyword(3) | task-oriented dialogue |
| Keyword(4) | spontaneous spoken dialogue |
| Keyword(5) | prosodic features |
| 1st Author's Name | Hanae Koiso |
| 1st Author's Affiliation | Chiba University() |
| 2nd Author's Name | Yasuo Horiuchi |
| 2nd Author's Affiliation | Chiba University |
| 3rd Author's Name | Syun Tutiya |
| 3rd Author's Affiliation | Chiba University |
| 4th Author's Name | Akira Ichikawa |
| 4th Author's Affiliation | Chiba University |
| Date | 1996/3/21 |
| Paper # | NLC95-72 |
| Volume (vol) | vol.95 |
| Number (no) | 600 |
| Page | pp.pp.- |
| #Pages | 6 |
| Date of Issue | |