Nakatani, C. and J. Hirschberg (1994). A corpus-based study of repair cues in spontaneous speech, Journal of the Acoustical Society of America, Vol. 95, No. 3, pp. 16031616.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....speech. Hindle [1] suggested that an acoustic edit signal serves as a cue that fluent speech has been interrupted. Although no evidence for a single such cue has been found, several corpus studies have found that combinations of cues can be used to identify disfluencies with reasonable success [2, 3, 4, 5]. The editing phase consists of a spoken cue phrase like filled pauses (such as uh) discourse markers (such as you know, I mean) or explicit editing terms (such as oops) Prosody Word LM POS LM Speech Human transcription IP hypotheses Repetition pattern LM Knowledge rules Fi n a lo ....

C. Nakatani and J. Hirschberg, "A corpus-based study of repair cues in spontaneous speech," JASA, pp. 1603--1616, 1994.

....and lower recall, many more words are left in the sentence to be parsed. Thus one finds more nonterminal constituents in the machine parses than in the gold parses and the precision is lower than the recall. 4 Previous research While there is a significant body of work on finding edit positions [1,9,10,13,17,18], it is di#cult to make meaningful comparisons between the various research e#orts as they di#er in (a) the corpora used for training and testing, b) the information available to the edit detector, and (c) the evaluation metrics used. For example, 13] uses a subsection of the ATIS corpus, takes ....

....work on finding edit positions [1,9,10,13,17,18] it is di#cult to make meaningful comparisons between the various research e#orts as they di#er in (a) the corpora used for training and testing, b) the information available to the edit detector, and (c) the evaluation metrics used. For example, [13] uses a subsection of the ATIS corpus, takes as input the actual speech signal (and thus has access to silence duration but not to words) and uses as its evaluation metric the percentage of time the program identifies the start of the interregnum (see Section 2.2) On the other hand, 9,10] use ....

Nakatani, C. H. and Hirschberg, J. A corpus-based study of repair cues in spontaneous speech. Journal of the Acoustical Society of America 953 (1994), 1603--1616.

.... Much work has been done on both detecting boundary tones (e.g. Wang and Hirschberg, 1992; Wightman and Ostendorf, 1994; Stolcke and Shriberg, 1996a; Kompe et al. 1994; Mast et al. 1996) and on speech repair detection and correction (e.g. Hindle, 1983; Bear, Dowding, and Shriberg, 1992; Nakatani and Hirschberg, 1994; Heeman and Allen, 1994; Stolcke and Shriberg, 1996b) This work has focused on one of the issues in isolation of the other. However, these two issues are intertwined. Cues such as the presence of silence, final syllable lengthening, and presence of filled pauses tend to mark both events. Even ....

....techniques that are used. Bear et al. 1992) used a simple pattern matching approach on ATIS word transcriptions. They exclude all turns that have a repair that just consists of a filled pause or word fragment. On this subset they obtained a correction recall rate of 43 and a precision of 50 . Nakatani and Hirschberg (1994) examined how speech repairs can be detected using a variety of information, including acoustic, presence of word matchings, and POS 4 Silence has a bigger effect on detection and correction if boundary tones are modeled. tags. Using these clues they were able to train a decision tree which ....

Nakatani, C. H. and J. Hirschberg. 1994. A corpus-based study of repair cues in spontaneous speech. Journal of the Acoustical Society of America, 95(3):1603--1616.

....discarding reparanda, that is, those parts of disfluencies which need to be expunged before a fluent sequence can be reconstructed. Much attention has been devoted to specifying the distinctive acoustic or structural features of disfluencies which would reliably trigger an accurate editing process [3, 5, 12, 14]. We are currently investigating the human perceptual solution to this problem. We propose that people do not mentally transcribe and then expunge disfluent speech. Instead, they fail to recognize the acoustic material of disfluencies as words or they recognize it with so much delay that portions ....

C Nakatani and J Hirschberg. A corpus-based study of repair cues in spontaneous speech. Journal of the Acoustical Society of America, 95:1603--1616, 1994.

.... intensity are perceived and processed in continuous speech is not advanced enough to be applicable in an ASR system [Vaissi ere, 88] 7 ffl as ASR systems aim at spontaneous speech recognition, and despite the growing number of prosodic studies on the subject [Ward, 91, Shriberg, 92, Blaauw, 94, Nakatani, 94] it is rather complicated to model prosodic knowledge (frequent hesitations, repairs, emphasis, agrammatical sentences, etc. ffl above all, the basic problem of prosodic analysis is that prosodic parameters are governed at the same time by several organizational levels which may become ....

C. H. Nakatani and J. Hirschberg. A corpus-based study of repair cues in spontaneous speech. Acoustical Society of America, 95(3):1603--1616, March 1994.

....prosodic features such as pause, duration, and pitch. Hidden events are thus suitable candidates for the kind of hidden structure needed to leverage prosody as a knowledge source for word recognition. Prosodic cues have been studied mainly for the purpose of automatic detection of disfluencies [10, 12] and sentence boundaries [8] The correlation between hidden events and word cues has likewise been exploited, for detecting both sentence boundaries [14, 8] and disfluencies [2, 6, among others] although recent work has also shown that speech language models can be improved by incorporating ....

C. H. Nakatani and J. Hirschberg. A corpus-based study of repair cues in spontaneous speech. Journal of the Acoustical Society of America, 95(3):1603--1616, 1994.

....becausenon error hesitations (filled pausesand repetitions) are suppressedin human computer dialog with a push to talk mechanism for speech input, errors make up a larger proportion of total disfluencies. Various researchers have described cutoffs as abrupt, showing some form of laryngealization [1, 14, 11]. In an analysis of cutoffs in the ATIS data conducted by Madelaine Plauche, we found that a typical form of laryngealization in such casescorrespondsto creaky voice on the last 20 50 ms of the cut off words. However, it is not the case that all cutoffs are sudden, or that word cutoffs always ....

Nakatani, C. H. and Hirschberg, J. 1994. A corpus-based study of repair cues in spontaneous speech. Journal Acoustical Society of America, 95(3), 1603--1616.

....of disfluencies could also benefit higher level modeling, for example, the automatic segmentation of speech into sentences [11] and the modeling of discourse or topic structure [13] 1.2. Why use prosody Various approaches to automatic disfluency detection have been proposed in past work [8, 1, 7, 4]. These studies have focused on task oriented dialog and have used a combination of lexical and prosodic features. Results have showna heavy reliance on lexical information, although prosodic information was also useful when constrained by the lexical information. Such approachesare limited, ....

C. H. Nakatani and J. Hirschberg. A corpus-based study of repair cues in spontaneous speech. Journal of the Acoustical Society of America, 95(3):1603--1616, 1994.

....strings or matching syntactic categories on either side of the editing signal and expunges them from further analysis. However, no clear definition of the editing signal has ever been proposed, and, like others, the model depends on successful prior recognition of words. Nakatani and Hirschberg [19, 20] propose a model of repair detection which takes into account speech based cues, but still relies heavily on textual cues, such as the presence of fragments and lexical and part of speech information. Research on human perception, however, suggests that prosodic, rather than syntactic or semantic ....

....part of the utterance that is in effect replaced by the repair) in the editing phase (from the end of the reparandum to the start of the repair) and in the repair itself, where speech resumes after the interruption. Glottalisation at the end of the reparandum has been suggested as a possible cue [19, 20, 3], particularly in vowel final fragments. Silent pause at the interruption point is seen as another indicator (see, for example [4] on the problem of how to define a pause) though (pre)pausal lengthening may provide a cue in a similar way, without any actual silence occurring [8] Filled pauses ....

[Article contains additional citation context not shown here]

C Nakatani and J Hirschberg. A corpus-based study of repair cues in spontaneous speech. Journal Of The Acoustical Society Of America, 95:1603--1616, 1994.

....read speech. DFs are one of many potential factors contributing to the relatively poor performance of state ofthe art recognizers on this type of speech, e.g. as found in the Switchboard [2] corpus. Past work on disfluent speech has focused on disfluency detection, using either acoustic features [7, 6] or recognized word sequences [1, 3] Our goal in this work is to develop a statistical language model (LM) that can be used for speech decoding or rescoring, and that improves upon standard LMs by explicitly modeling the most frequent DF types. The main reason to expect that DF modeling can ....

C. H. Nakatani and J. Hirschberg. A corpus-based study of repair cues in spontaneous speech. Journal of the Acoustical Society of America, 95(3):1603--1616, 1994.

....communication and speech processing by machine. Although historically disfluencies have been viewed as noisy events, and have received relatively little attention, a more recent focus on spontaneous speech has directed increased interest to disfluencies in both theoretical and applied fields [1,4,6,8,9,10,11]. The goal of the present work is to illustrate that disfluencies are not noise but rather show systematic distributions in various dimensions. This paper summarizes results from a large descriptive study aimed at revealing and modeling trends in the distribution and form of disfluencies in ....

Nakatani, C.H. and Hirschberg, J. "A corpus-based study of repair cues in spontaneous speech," Journal of the Acoustical Society of America, 95(3), 1603-1616, 1994.

....94] PhD theses. These approaches look for prosodic clues which indicate interruption points, or word patterns which help indicate how far the reparanda extends backwards. Non syntactic methods for detecting and removing disfluencies can be clasified as: pre recognition : O Shaughnessy 92] Nakatani Hirschberg 94] study the speech signal to find clues to disfluencies. pre parsing and post recognition : Heeman Allen 94] assume a perfect transcription, but detect disfluencies using word pattern matching and part of speech information provided by a POS tagger, but without using a parser. post parsing : ....

Nakatani, C.H., & Hirschberg,J., "A corpus-based study of repair cues in spontaneous speech", Journal of the Acoustical Society of America, 95(3), 1994.

.... Human listeners are reasonably accurate in transcribing fluent speech but find it difficult to transcribe disfluencies [10] In contrast, automatic speech recognition systems have considerable difficulty spotting and excising disfluencies despite the distinctive acoustic or structural features [3, 6, 11] of these very common phenomena. We report on the human abilities which make disfluencies evanescent. The portion of a disfluent utterance which must be expunged to make fair copy is called the reparandum. Though words in reparanda are processed [4] they may not be correctly identified by normal ....

C. Nakatani & J. Hirschberg. "A corpus-based study of repair cues in spontaneous speech". JASA, 95:1603-

No context found.

C. Nakatani and J. Hirschberg. A corpus-based study of repair cues in spontaneous speech. JASA, 1994.

No context found.

Nakatani, C. and J. Hirschberg (1994). A corpus-based study of repair cues in spontaneous speech, Journal of the Acoustical Society of America, Vol. 95, No. 3, pp. 16031616.

No context found.

C. Nakatani and J. Hirschberg. A corpus-based study of repair cues in spontaneous speech. Journal of the Acoustical Society of America, pages 1603--1616, 1994.

No context found.

C. Nakatani, J. Hirschberg, 1994. A Corpus-based Study of Repair Cues in Spontaneous Speech. Journal of the Acoustical Society of America, pp. 1603-1616.

No context found.

C. Nakatani and J. Hirschberg, "A corpus-based study of repair cues in spontaneous speech," JASA, pp. 1603--1616, 1994.

No context found.

C. H. Nakatani and J. Hirschberg, "A corpus-based study of repair cues in spontaneous speech," J. Acoust. Soc. Am., vol. 95, no. 3, pp. 1603-1616, 1994.

No context found.

Nakatani, Christine H. & Julia Hirschberg. 1994. A corpus-based study of repair cues in spontaneous speech. Journal of the Acoustical Society of America, vol. 954, pp. 1603--1616.

No context found.

Nakatani, C. and Hirschberg, J. (1994). A Corpus-based Study of Repair Cues in Spontaneous Speech. Journal of the Acoustical Society of America, 95(3):1603 -- 1616.

No context found.

C. Nakatani and J. Hirschberg, "A corpus-based study of repair cues in spontaneous speech," JASA, pp. 1603--1616, 1994.

No context found.

Christine H. Nakatani and Julia Hirschberg. A corpus-based study of repair cues in spontaneous speech. J. Acoust. Soc. Am, 95(3):1603--1616, 1994.

No context found.

C Nakatani and J Hirschberg. A corpus-based study of repair cues in spontaneous speech. Journal Of The Acoustical Society Of America, 95:1603#1616, 1994.

No context found.

C Nakatani and J Hirschberg. A corpus-based study of repair cues in spontaneous speech. Journal of the Acoustical Society of America, 95:1603#1616, 1994.

No context found.

Nakatani, C. (1994) \A corpus-based Study of Repair Cues in Spontaneous Speech", J. Acoust. Soc. Am. 95 (3) March 1994.

No context found.

C. Nakatani and J. Hirschberg. (in press). A corpus-based study of repair cues in spontaneous speech. In Journal of the Acoustical Society of America.

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