Using Prosody for Automatic Sentence Segmentation of Multi-party Meetings

Kolář, Jáchym; Shriberg, Elizabeth; Liu, Yang

doi:10.1007/11846406_79

Cited by 18 publications

(15 citation statements)

References 16 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…BoosTexter was initially designed for the task of text categorization, employment of this method for tasks related to DA segmentation was firstly presented in [9,19]. The method combines weak classifiers having a basic form of one-level decision trees (stumps) using confidence-rated predictions.…”

Section: Boosting-based Models (Boostexter)mentioning

confidence: 99%

A Comparison of Language Models for Dialog Act Segmentation of Meeting Transcripts

Kolář

Text, Speech and Dialogue

Self Cite

View full text Add to dashboard Cite

Abstract. This paper compares language modeling techniques for dialog act segmentation of multiparty meetings. The evaluation is twofold; we search for a convenient representation of textual information and an efficient modeling approach. The textual features capture word identities, parts-of-speech, and automatically induced classes. The models under examination include hidden event language models, maximum entropy, and BoosTexter. All presented methods are tested using both human-generated reference transcripts and automatic transcripts obtained from a state-of-the-art speech recognizer.

show abstract

Section: Boosting-based Models (Boostexter)mentioning

confidence: 99%

A Comparison of Language Models for Dialog Act Segmentation of Meeting Transcripts

Kolář

Text, Speech and Dialogue

Self Cite

View full text Add to dashboard Cite

show abstract

“…Several different approaches utilizing both textual (lexical or syntactic) and acoustic (prosodic) information have been proposed. The proposed techniques include hidden Markov models (HMMs) (Shriberg et al, 2000;Kim and Woodland, 2003), multilayer perceptrons (Warnke et al, 1997;Srivastava and Kubala, 2003), maximum entropy (Huang and Zweig, 2002;Liu et al, 2004), conditional random fields (Liu et al, 2005;Zimmermann, 2009), support vector machines (Akita et al, 2006;Magimai-Doss et al, 2007), and adaptive boosting (Zimmermann et al, 2006;Kolář et al, 2006b). Syntactic information has been used in (Roark et al, 2006;Favre et al, 2008).…”

Section: Related Workmentioning

confidence: 99%

Speaker adaptation of language and prosodic models for automatic dialog act segmentation of speech

Kolář¹,

Liu²,

Shriberg³

2010

Speech Communication

Self Cite

View full text Add to dashboard Cite

Speaker-dependent modeling has a long history in speech recognition, but has received less attention in speech understanding. This study explores speaker-specific modeling for the task of automatic segmentation of speech into dialog acts (DAs), using a linear combination of speaker-dependent and speaker-independent language and prosodic models. Data come from 20 frequent speakers in the ICSI meeting corpus; adaptation data per speaker ranges from 5k to 115k words. We compare performance for both reference transcripts and automatic speech recognition output. We find that: (1) speaker adaptation in this domain results both in a significant overall improvement and in improvements for many individual speakers, (2) the magnitude of improvement for individual speakers does not depend on the amount of adaptation data, and (3) language and prosodic models differ both in degree of improvement, and in relative benefit for specific DA classes. These results suggest important future directions for speaker-specific modeling in spoken language understanding tasks.

show abstract

“…In general, the information sources we use for Czech and English are similar. See [7] for a more detailed description of the features, along with an analysis of their contribution to sentence segmentation of Czech, and [8] for more information about English.…”

Section: Knowledge Sourcesmentioning

confidence: 99%

Automatic sentence boundary detection in conversational speech: A cross-lingual evaluation on English and Czech

Kolář

Liu²

2010

2010 IEEE International Conference on Acoustics, Speech and Signal Processing

Self Cite

View full text Add to dashboard Cite

Automatic sentence segmentation of speech is important for enriching speech recognition output and aiding downstream language processing. This paper focuses on automatic sentence segmentation of speech in two different languages -English and Czech. For this task, we compare and combine three statistical models -HMM, maximum entropy, and a boosting-based model BoosTexter. All these approaches rely on both textual and prosodic information. We evaluate these methods on a corpus of multiparty meetings in English, and on a corpus of broadcast conversations in Czech, using both manual and speech recognition transcripts. The experiments show that superior results are achieved when all the three models are combined via posterior probability interpolation. We observe differences in terms of model performance between English and Czech, as well as the feature usage difference in prosodic models between the two languages. Overall, the analysis is important for porting sentence segmentation approaches from one language to another.

show abstract

Using Prosody for Automatic Sentence Segmentation of Multi-party Meetings

Cited by 18 publications

References 16 publications

A Comparison of Language Models for Dialog Act Segmentation of Meeting Transcripts

A Comparison of Language Models for Dialog Act Segmentation of Meeting Transcripts

Speaker adaptation of language and prosodic models for automatic dialog act segmentation of speech

Automatic sentence boundary detection in conversational speech: A cross-lingual evaluation on English and Czech

Contact Info

Product

Resources

About