PCFGs with Syntactic and Prosodic Indicators of Speech Repairs

A grammatical method of combining two kinds of speech repair cues is presented. One cue, prosodic disjuncture, is detected by a decision tree-based ensemble classifier that uses acoustic cues to identify where normal prosody seems to be interrupted (Lickley, 1996). The other cue, syntactic parallelism, codifies the expectation that repairs continue a syntactic category that was left unfinished in the reparandum (Levelt, 1983). The two cues are combined in a Treebank PCFG whose states are split using a few simple tree transformations. Parsing performance on the Switchboard and Fisher corpora suggests that these two cues help to locate speech repairs in a synergistic way.

• Publication year

  2006

• Venue

  Annual Meeting of the Association for Computational Linguistics

• Publication date

  2006-07-17

• Fields of study

  Linguistics, Computer Science

• Identifiers
  DOI 10.3115/1220175.1220196
• External record

  Open on Semantic Scholar

• Source metadata

  Semantic Scholar

• No claims are published for this paper.

• No concepts are published for this paper.

• Effective Use of Prosody in Parsing Conversational Speech

  2005cited by this paper
• Parsing Partially Bracketed Input

  2005cited by this paper
• 2005 Johns Hopkins Summer Workshop Final Report on Parsing and Spoken Structural Event Detection

  2005cited by this paper
• Exploring Features for Identifying Edited Regions in Disfluent Sentences

  2005cited by this paper
• Modeling prosodic feature sequences for speaker recognition

  2005cited by this paper
• Parsing and its applications for conversational speech

  2005cited by this paper
• Sentence-Internal Prosody Does not Help Parsing the Way Punctuation Does

  2004cited by this paper
• Robust garden path parsing

  2004cited by this paper
• An Improved Model for Recognizing Disfluencies in Conversational Speech

  2004cited by this paper
• A TAG-based noisy-channel model of speech repairs

  2004influential reference
• Structural event detection for rich transcription of speech

  2004cited by this paper
• How to repair speech repairs in an end-to-end system

  2001cited by this paper
• Edit Detection and Parsing for Transcribed Speech

  2001cited by this paper
• A prosodically labeled database of spontaneous speech

  2001cited by this paper
• Enriching the Knowledge Sources Used in a Maximum Entropy Part-of-Speech Tagger

  2000cited by this paper
• Processing Self Corrections in a speech to speech system

  2000cited by this paper
• A Maximum-Entropy-Inspired Parser

  2000cited by this paper
• Speech repains, intonational phrases, and discourse markers: modeling speakers’ utterances in spoken dialogue

  1999cited by this paper
• A Syntactic Framework for Speech Repairs and Other Disruptions

  1999cited by this paper
• The Syntax of Disfluency in Spontaneous Spoken Language

  1998cited by this paper
• Modeling dynamic prosodic variation for speaker verification

  1998cited by this paper
• Speech Repairs, Intonational Boundaries and Discourse Markers: Modeling Speakers' Utterances in Spoken Dialog

  1997cited by this paper
• A Maximum Entropy Model for Part-Of-Speech Tagging

  1996cited by this paper
• Juncture cues to disfluency

  1996cited by this paper
• Bagging Predictors

  1996influential reference
• Statistical language modeling for speech disfluencies

  1996cited by this paper
• Preliminaries to a Theory of Speech Disfluencies

  1994influential reference
• A corpus-based study of repair cues in spontaneous speech.

  1994cited by this paper
• The Penn Treebank: Annotating Predicate Argument Structure

  1994cited by this paper
• Building a Large Annotated Corpus of English: The Penn Treebank

  1993cited by this paper
• TOBI: a standard for labeling English prosody

  1992cited by this paper
• SWITCHBOARD: telephone speech corpus for research and development

  1992influential reference
• The Use of Prosody in Syntactic Disambiguation

  1991cited by this paper
• Deterministic Parsing of Syntactic Non-fluencies

  1983cited by this paper
• Meta-rules as a Basis for Processing Ill-Formed input

  1983cited by this paper
• Monitoring and self-repair in speech.

  1983cited by this paper

• Assessing the Use of Prosody in Constituency Parsing of Imperfect Transcripts

  2021cites this paper
• Prosodic segmentation for parsing spoken dialogue

  2021cites this paper
• Neural Constituency Parsing of Speech Transcripts

  2019cites this paper
• On the Role of Style in Parsing Speech with Neural Models

  2019cites this paper
• Parsing Speech: a Neural Approach to Integrating Lexical and Acoustic-Prosodic Information

  2017cites this paper
• Joint Modeling of Text and Acoustic-Prosodic Cues for Neural Parsing

  2017cites this paper
• Grammatical Knowledge is Fundamentally Probabilistic

  2015cites this paper
• Unediting: Detecting Disfluencies Without Careful Transcripts

  2015cites this paper
• Multi-domain disfluency and repair detection

  2014cites this paper
• Modelling incremental self-repair processing in dialogue

  2014cites this paper
• Rethinking Deep and Surface : towards a comprehensive model of anaphoric processes in dialogue

  2013cites this paper
• A sequential repetition model for improved disfluency detection

  2013cites this paper
• Word Buffering Models for Improved Speech Repair Parsing

  2009cites this paper
• Parsing Speech Repair without Specialized Grammar Symbols

  2009cites this paper
• Improved Syntactic Models for Parsing Speech with Repairs

  2009influential citation
• Knowledge integration into language models: a random forest approach

  2009cites this paper
• A Syntactic Time-Series Model for Parsing Fluent and Disfluent Speech

  2008influential citation
• A Unified Syntactic Model for Parsing Fluent and Disfluent Speech

  2008cites this paper
• Exploiting prosodic breaks in language modeling with random forests

  2008influential citation
• DEPENDENCY PARSING OF SPOKEN SWEDISH

  2008cites this paper
• Exploiting prosody for PCFGs with latent annotations

  2007cites this paper
• Incorporating discourse and syntactic dependencies into probabilistic models for summarization of multiparty speech

  2007cites this paper