Grosz, B., & Hirschberg, J.: Some intonational characteristics of discourse structure. Proceedings of 2 nd of International Conference on Spoken Language Processing, 1, (1992) 429--432.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....present in speech, such as intonation and the use of pauses, are not generally transcribed. Hirschberg and Grosz studied the relationship between intonation and discourse structure and found correlations between intonational features and annotator s labeling of discourse structure [Hirschberg and Grosz, 1992, Grosz and Hirschberg, 1992] Hirschberg and Nakatani showed that annotators more consistently segment discourse using speech in conjunction with transcripts than using text alone and that annotators can reliably segment both spontaneous and planned speech [Hirschberg and Nakatani, 1996] ....

....such as intonation and the use of pauses, are not generally transcribed. Hirschberg and Grosz studied the relationship between intonation and discourse structure and found correlations between intonational features and annotator s labeling of discourse structure [Hirschberg and Grosz, 1992, Grosz and Hirschberg, 1992] Hirschberg and Nakatani showed that annotators more consistently segment discourse using speech in conjunction with transcripts than using text alone and that annotators can reliably segment both spontaneous and planned speech [Hirschberg and Nakatani, 1996] Passoneau and Litman studied ....

[Article contains additional citation context not shown here]

Grosz, B. J. and Hirschberg, J. (1992). Some intonational characteristics of discourse structure. In International Conference on Spoken Language Processing, pages 429--432.

....into four classes: silence, clean speech, noisy speech, music environmental sounds. In order to ensure coherent segments, we smooth the result using a duration dependent Viterbi decoder. Second, we analyze the discourse structure, for significant phrases, using the acoustic correlates of prosody [7][12] in conjunction with an SVM classifier. The two modes of analysis are then combined and the speech segments are ranked in order of significance. We formulate the problem of skim generation in terms of utility maximization with constraints. We model the skim comprehensibility in terms of audio ....

....duration distribution of each class) to find the maximum likelihood class path. 6.3 Detecting significant phrases In this section, we shall summarize our work on detecting segment beginnings (SBEG s) in speech. These are important as they serve as the introduction of new topic in the discourse [7], Detecting these discourse boundaries is different from determining emphasized portions of speech [2] since these emphasized portions can occur anywhere in the discourse (including SBEG s) There has been much work in the computational linguistics community [8] 7] 12] to determine the acoustic ....

[Article contains additional citation context not shown here]

J. Hirschberg, B. Groz, Some Intonational Characteristics of Discourse Structure, Proc. ICSLP 1992.

....example continuation, elaboration, contrast, etc. It has been shown that the prosodic realization of spoken discourse reflects to some degree the inherent semantic structure of the discourse. Both pause duration and the pitch range correlate strongly with the coarse topic structure of a discourse [1, 2]. Whereas most of these studies are based on relatively short stretches of speech from few speakers we are aiming at a model of discourse structure that can be evaluated on a larger speech database. Furthermore, our goal is to establish a relation between the discourse structure and its prosodic ....

B. Grosz and C. L. Sidner, "Some intonational characteristics of discourse structure," in Proceedings of International Conference on Spoken Language Processing, Banff, Canada, 1992, pp. 429--432.

....for example continuation, elaboration, contrast, etc. It has been shown that the prosodic realization of spoken discourse reflects to some degree the inherent semantic structure of the discourse. Grosz Hirschberg show that pause duration correlates with the rough topic structure of a discourse [1]. Another parameter under consideration is pitch range [2] But most of these studies are based on relatively short stretches of speech from few speakers. In view of recent efforts in corpus based linguistics, it appears that research on prosody, and particularly on discourse prosody, yields more ....

Grosz, B.J. & Hirschberg, J. (1992). Some intonational characteristics of discourse structure. Proceedings of the 2 nd International Conference on Spoken Language Processing, Banff, Canada: 429-432.

....boundaries are marked by a greater degree of Final Lengthening and Silent Interval duration than those associated with PPhs. The beginnings of PU s are also characterized acoustically by a high F0 level on the first accented word. This seems to be a general characteristic of discourse intonation [7]; however, we have not as yet made any detailed study of this and other left edge boundary markers in our data. 2. IMPLEMENTING THE PROSODIC STRUCTURE Using the information obtained from the referent tracker and the prosodic parser, we are currently involved in developing a rule component for ....

Grosz, B. & Hirschberg, J. "Some intonational characteristics of discourse structure," Proceedings ICSLP92, Vol. 1:429-432, 1992.

....between discourse segments. In some cases, the role of these relations was considered to fall outside the scope of a study [Flammia and Zue, 1995] in other cases, judgments were made with respect to a taxonomy of very few intention based relations (usually dominance and satisfaction precedence) [Grosz and Hirschberg, 1992, Nakatani et al. 1995, Hirschberg and Litman, 1987, Passonneau and Litman, 1997, Carletta et al. 1997] And in the only case in which a rich taxonomy of 29 relations was used [Moser and Moore, 1997] the corpus was small and specific to a very restricted genre: written interactions between a ....

Barbara J. Grosz and Julia Hirschberg. Some intonational characteristics of discourse structure. In Proceedings of the International Conference on Spoken Language Processing, 1992.

....intentions of the utterances, such as the how the current utterance relates to prior discourse. We have identified four types of discourse cues. The first type is perceptible silence, or pauses, observed at the end of an utterance, which has been found to correlate with discourse boundaries (Grosz and Hirschberg, 1992;Passonneau and Litman, 1993; Swerts, 1997) We believe that in the context of initiative modeling, silence at the end of an utterance may suggest that the speaker has nothing more to say in the current turn and intends to give up his task dialogue initiative. For instance, in the following ....

Grosz, B. and J. Hirschberg: 1992, `Some Intonational Characteristics of Discourse Structure'. In: Proceedings of the International Conference on Spoken Language Processing. pp. 429--432.

....dialogues. There have been a number of studies on this issue. Hakoda and Sato [11] claimed that when one read aloud written texts, the syntactic structure of each sentence is reflected in the prosodic parameters; onset, peak, and final F0 values, of each intonational phrase. Grosz and Hirschberg [8] analyzed AP news stories spoken by a newscaster and confirmed that there was a correlation between the discourse features such as the discourse segment boundaries and the prosodic features: the F0 range and pause insertion. Fujisaki [7] also investigated the narration of professional announcers ....

Grosz, B.; Hirschberg, J.: Some intonational characteristics of discourse structure. Proceedings of International Conference on Spoken Language Processing: 429-432 (1992).

....pitch accents contain information about the status of discourse referents, phrasal tones about the relatedness of intermediate phrases, and boundary tones about whether the phrase is forward looking or not. Intonation can also give information about discourse structure. Grosz and Hirschberg (Grosz and Hirschberg, 1992; Hirschberg and Grosz, 1992) investigated whether subjects could segment a dialogue based on Grosz and Sidner s theory of how discourse structure relates to the intentional and attentional state of the participants (1986) As their corpus, they used AP news stories read by a professional ....

....the status of discourse referents, phrasal tones about the relatedness of intermediate phrases, and boundary tones about whether the phrase is forward looking or not. Intonation can also give information about discourse structure. Grosz and Hirschberg (Grosz and Hirschberg, 1992; Hirschberg and Grosz, 1992) investigated whether subjects could segment a dialogue based on Grosz and Sidner s theory of how discourse structure relates to the intentional and attentional state of the participants (1986) As their corpus, they used AP news stories read by a professional newscaster. They had a group of ....

[Article contains additional citation context not shown here]

Grosz, Barbara and Julia Hirschberg. 1992. Some intonational characteristics of discourse structure. In Proceedings of the 2nd International Conference on Spoken Language Processing (ICSLP-92), pages 429--432, October.

....derived theories already achieve optimal performance with respect to the examined features in this less noisy subcorpus, and or that the automatically derived theory for this subcorpus was based on a smaller training set than used in the previous subcorpus. Related Work in Discourse Analysis Grosz and Hirschberg (1992) used the system cart (Brieman et al. 1984) to construct decision trees for classifying aspects of discourse structure from intonational feature values. Siegel (in press) was the first to apply machine learning to cue phrases. He developed a genetic learning algorithm to induce decision trees ....

Grosz, B., and Hirschberg, J. 1992. Some intonational characteristics of discourse structure. In Proc.

....may be able to infer the completion of a SharedPlan, and thus the corresponding segment, in combination with information from other sources. For example, suppose an OCP has some reason to expect the end of a segment based on a linguistic signal such as an intonational feature (e.g. as described by Grosz and Hirschberg (1992)) If additionally the OCP is able to ascribe the various mental attitudes missing from the SharedPlan that corresponds to that segment, then the OCP has further evidence for the segment boundary. Understanding utterances within discourse segments Under Grosz and Sidner s theory, each utterance ....

Grosz, B. and J. Hirschberg. 1992. Some intonational characteristics of discourse structure. In Proceedings of the International Conference on Spoken Language Processing, Banff, Alberta, Canada, October 12-16.

.... claimed to constrain and be constrained by disparate phenomena: cue phrases (Hirschberg and Litman, 1993; Grosz and Sidner, 1986; Reichman, 1985; Cohen, 1984) lexical cohesion (Morris and Hirst, 1991) plans and intentions (Carberry, 1990; Litman and Allen, 1990; Grosz and Sidner, 1986) prosody (Grosz and Hirschberg, 1992; Hirschberg and Grosz, 1992; Hirschberg and Pierrehumbert, 1986) reference (Webber, 1991; Grosz and Sidner, 1986; Linde, 1979) and tense (Webber, 1988; Hwang and Schubert, 1992; Song and Cohen, 1991) However, there is weak consensus on the nature of segments and the criteria for recognizing or ....

.... disparate phenomena: cue phrases (Hirschberg and Litman, 1993; Grosz and Sidner, 1986; Reichman, 1985; Cohen, 1984) lexical cohesion (Morris and Hirst, 1991) plans and intentions (Carberry, 1990; Litman and Allen, 1990; Grosz and Sidner, 1986) prosody (Grosz and Hirschberg, 1992; Hirschberg and Grosz, 1992; Hirschberg and Pierrehumbert, 1986) reference (Webber, 1991; Grosz and Sidner, 1986; Linde, 1979) and tense (Webber, 1988; Hwang and Schubert, 1992; Song and Cohen, 1991) However, there is weak consensus on the nature of segments and the criteria for recognizing or generating them in a ....

[Article contains additional citation context not shown here]

B. Grosz and J. Hirschberg. 1992. Some intonational characteristics of discourse structure. In Proc. of the International Conference on Spoken Language Processing.

No context found.

Barbara Grosz and Julia Hirschberg. Some intonational characteristics of discourse structure. In Proceedings of ICSLP-92, Banff, October 1992.

....the discourse model was simplified in this system, many facets of discourse structure were hand coded, and variation in intonational features was limited. In research with Julia Hirschberg at AT T Bell Laboratories, we performed a pilot study of discourse structure and intonation based on G S 3 [52, 41]. This study, supported in part by a grant from NSF, examined both the global level (discourse segments and relations between them) and the local level (features of utterances within a discourse segment) of discourse. We found statistically significant correlations of aspects of pitch range, ....

Barbara Grosz and Julia Hirschberg. Some intonational characteristics of discourse structure. In Proc. of International Conference on Spoken Language Processing, Banff, Alberta, Canada, October 12-16 1992. to appear.

....language model. Currently, these same units are used for visual browsing and play back in the SCAN GUI. Better choice of boundaries for these paratones can be made using simple pausal duration information, for a given speaker; that is, longer pauses are reliably correlated with topic beginnings [2, 3, 4]. However, it is difficult to find topic boundaries reliably across speakers, due to differences in speaking rate. 2.2. Classifying Channel Conditions The intonational paratones classified by CART are then passed to a simple Gaussian mixture based classifier that divides them into wide band or ....

Barbara Grosz and Julia Hirschberg. Some intonational characteristics of discourse structure. In Proceedings of the International Conference on Spoken Language Processing, Banff, October 1992. ICSLP.

....the same contour is used with an expanded pitch range and a faster rate to convey incredulity. Range variation can also convey differences in degree of speaker involvement , or communicate the topic structure of a text (Hirschberg and Pierrehumbert, 1986; Pierrehumbert and Hirschberg, 1990; Grosz and Hirschberg, 1992). Size of Overall Range In general, new topics are begun in a wider pitch range and end in a narrower range. Subsequent expansion of pitch range tends to be associated with a topic shift. So, for example, the structure of the text in Example (45) can be conveyed by range variation; range ....

....example, n f T150 changes TTS topline to 150 for the entire invocation of the program. Other discourse phenomena can be conveyed by variation of these range parameters. For example, parentheticals are generally uttered in a compressed pitch range and with a faster speaking rate than other phrases (Grosz and Hirschberg, 1992), as illustrated in Example (46) 46) n fT130 The crowd, n fT105 n r.7 which had been waiting for hours in the snow, n fT130 n r1 became unruly. Attributive tags (e.g. he said ) are also characterized by a compressed pitch range (Grosz and Hirschberg, 1992) as illustrated in Example (47) 47) ....

[Article contains additional citation context not shown here]

Grosz, B. and J. Hirschberg. 1992. Some intonational characteristics of discourse structure. Technical Memorandum TM-11222-92-1113-12, AT&T Bell Laboratories.

No context found.

Grosz, B., & Hirschberg, J.: Some intonational characteristics of discourse structure. Proceedings of 2 nd of International Conference on Spoken Language Processing, 1, (1992) 429--432.

No context found.

Grosz, Barbara and Ju lia H irschberg. 1992. Some intonational characteristics of d iscourse structu re . In Proceedings of 2nd of International Conference on Spoken Language Processing, 1992, Vol 1, pp. 429-432.

No context found.

Grosz, B., & Hirschberg, J., (1992) Some intonational characteristics of discourse structure. In Proceedings of 2nd of International Conference on Spoken Language Processing, 1992, Vol 1, pp. 429-432.

No context found.

Grosz, B., & Hirschberg, J.: Some intonational characteristics of discourse structure. Proceedings of 2 nd of International Conference on Spoken Language Processing, 1, (1992) 429--432.

No context found.

Grosz, B., & Hirschberg, J., (1992) Some intonational characteristics of discourse structure. In Proceedings of 2nd of International Conference on Spoken Language Processing, 1992, Vol 1, pp. 429-432.

No context found.

B. Grosz and J. Hirschberg, "Some intonational characteristics of discourse structure," Proc. 2nd International Conf. on Spoken Language Processing, vol.1, pp.429--432, Ban#, Canada, 1992.

No context found.

B. Grosz and J. Hirschberg. 1992. Some intonational characteristics of discourse structure. In Proc. of ICSLP-92, pages 429--432.

No context found.

Barbara J. Grosz and Julia B. Hirschberg. Some intonational characteristics of discourse structure. In ICSLP, 1992.

No context found.

Barbara Grosz and Julia Hirschberg. 1992. Some intonational characteristics of discourse structure. In Proceedings of the International Conference on Spoken Language Processing.

First 50 documents

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC