A Prosodic Analysis of Discourse Segments in Direction-Giving Monologues

A Prosodic Analysis of Discourse Segments in Direction- Giving Monologues Julia Hirschberg AT&T Laboratories, 2C-409 600 Mountain Avenue Murray Hill, NJ 07974 Christine H. Nakatani* Harvard University 33 Oxford Street Cambridge, MA 02138 Abstract This paper reports on corpus-based research into the relationship between intonational vari- ation and discourse structure. We examine the effects of speaking style (read versus sponta- neous) and of discourse segmentation method (text-alone versus text-and-speech) on the na- ture of this relationship. We also compare the acoustic-prosodic features of initial, me- dial, and final utterances in a discourse seg- ment. 1 Introduction This paper presents empirical support for the as- sumption long held by computational linguists, that intonation can provide valuable cues for discourse processing. The relationship between intonational variation and discourse structure has been explored in a new corpus of direction-giving monologues. We examine the effects of speaking style (read versus spontaneous) and of discourse segmentation method (text-alone versus text-and-speech) on the nature of this relationship. We also compare the acoustic- prosodic features of initial, medial, and final utter- ances in a discourse segment. A better understand- ing of the role of intonation in conveying discourse structure will enable improvements in the natural- ness of intonational variation in text-to-speech sys- tems as well as in algorithms for recognizing dis- course structure in speech-understanding systems. 2 Theoretical and Empirical Foundations It has long been assumed in computational lin- guistics that discourse structure plays an impor- tant role in Natural Language Understanding tasks such as identifying speaker intentions and resolving anaphoric reference. Previous research has found *The second author was partially supported by NSF Grants No. IRI-90-09018, No. IRI-93-08173, and No. CDA-94-01024 at Harvard University and by AT&T Bell Laboratories. that discourse structural information can be inferred from orthographic cues in text, such as paragraph- ing and punctuation; from linguistic cues in text or speech, such as cue PHI~.ASES 1 (Cohen, 1984; Reichman, 1985; Grosz and Sidner, 1986; Passon- neau and Litman, 1993; Passonneau and Litman, to appear) and other lexical cues (Hinkelman and Allen, 1989); from variation in referring expres- sions (Linde, 1979; Levy, 1984; Grosz and Sidner, 1986; Webber, 1988; Song and Cohen, 1991; Passon- neau and Litman, 1993), tense, and aspect (Schu- bert and Hwang, 1990; Song and Cohen, 1991); from knowledge of the domain, especially for task- oriented discourses (Grosz, 1978); and from speaker intentions (Carberry, 1990; Litman and Hirschberg, 1990; Lochbaum, 1994). Recent methods for auto- matic recognition of discourse structure from text have incorporated thesaurus-based and other in- formation retrieval techniques to identify changes in topic (Morris and Hirst, 1991; Yarowsky, 1991; Iwafiska et al., 1991; Hearst, 1994; Reynar, 1994). Parallel investigations on prosodic/acoustic cues to discourse structure have investigated the contri- butions of features such as pitch range, pausal du- ration, amplitude, speaking rate, and intonational contour to signaling topic change. Variation in pitch range has often been seen as conveying 'topic struc- ture' in discourse. Brown et al. (1980) found that subjects typically started new topics relatively high in …

Content truncated to 500 words