ORPP logo
Image from Google Jackets

Rhapsodie : A Prosodic and Syntactic Treebank for Spoken French.

By: Contributor(s): Material type: TextTextSeries: Studies in Corpus Linguistics SeriesPublisher: Amsterdam/Philadelphia : John Benjamins Publishing Company, 2019Copyright date: ©2019Edition: 1st edDescription: 1 online resource (414 pages)Content type:
  • text
Media type:
  • computer
Carrier type:
  • online resource
ISBN:
  • 9789027262929
Subject(s): Genre/Form: Additional physical formats: Print version:: RhapsodieDDC classification:
  • 448.007
LOC classification:
  • PC2066
Online resources:
Contents:
Intro -- Rhapsodie -- Editorial page -- Title page -- Copyright page -- Table of contents -- Acknowledgments -- Preface -- Introduction -- 1. Collecting data for the Rhapsodie treebank -- 1. Introduction -- 2. Purpose, method and context -- 2.1 Theoretical approach -- 2.2 Previous projects -- 3. Rhapsodie sampling -- 3.1 Rhapsodie corpus design: General principles -- 3.2 Gathering data: External and internal sources -- 4. Questions to be answered -- 4.1 Legal and ethical issues -- 4.2 Metadata -- 5. Conclusions -- 2. Orthographic and phonetic transcriptions of Rhapsodie recording -- 1. Introduction -- 2. From sound to orthographic transcription -- 2.1 The myth of the copyist -- 2.2. "Transcription as theory" -- 2.3 What to transcribe? -- 3. The three leading principles of orthographic transcription in Rhapsodie -- 3.1 Standard spelling -- 3.2 Punctuation and silent pauses -- 3.3 Handling of features that are specific to speech -- 3.4 Transcription conventions -- 3.5 Availability of orthographic transcription in two formats -- 4. Phonetic transcription and segmentation -- 4.1 Macro-segmentation at utterance level -- 4.2 Grapheme-to-phoneme conversion -- 4.3 Phone segmentation -- 4.4 Creation of a syllabic tier -- 5. Specific processing of multi-speaker recordings -- 6. Quality and productivity -- 7. Conclusion -- 3. Syntactic annotation of the Rhapsodie corpus -- 1. Annotating three mechanisms of cohesion and two levels of analysis -- 1.1 Microsyntax: Government and listing -- 1.1.1 Government -- 1.1.2 Listing -- 1.2 Macrosyntax -- 1.3 Government Units and Illocutionary Units -- 2. The advantages of annotating a complex structure -- 2.1 A rich syntactic treebank -- 2.2 Discourse markers, reformulations and disfluencies as part of the syntactic structure of spoken language -- 2.3 A syntactic annotation without sentences -- 3. Conclusion.
4. Microsyntactic annotation -- 1. Introduction -- 2. Morpho-syntactic Analysis -- 2.1 Word-forms and lexemes -- 2.2 Parts of speech -- 2.3 Morpho-syntactic features -- 3. Microsyntactic structure -- 3.1 Government -- 3.2 Dependency -- 3.3 Government unit -- 3.4 Choice of the head -- 4. Syntactic functions -- 5. Some complex cases -- 5.1 Extraction and qu-words -- 5.2 Negation -- 6. Phrase structure -- 7. Conclusion -- 5. The annotation of list structures -- 1. Introduction -- 2. Structure of lists -- 2.1 Layers -- 2.2 Conjuncts -- 2.3 Junctors -- 2.4 Paradigmatizing adverbs -- 2.5. List completers -- 2.6 Dependency and inherited dependency -- 2.7 Paradigmatic links -- 2.8 Junction links -- 3. Resolution of complex cases of coordination -- 3.1 Scope ambiguity -- 3.2 Gapping and non-constituent coordination -- 3.3 Embedded coordinations -- 3.4 Correlative structures -- 3.5 Junction without list -- 4. Types of lists -- 4.1 Relational coordination (para_coord) -- 4.2 Hypernymic coordination (para_hyper) -- 4.3 Intensification (para_intens) -- 4.4 Disfluency (para_disfl) -- 4.5 Reformulation (para_reform) -- 4.6 Double formulation (para_dform) -- 4.7 Negotiating formulation (para_negot) -- 5. Conclusion -- 6. Macrosyntactic annotation -- 1. Introduction -- 1.1 Macrosyntactic traditions -- 1.2 Rhapsodie's approach to macrosyntax -- 2 Rhapsodie's macrosyntactic annotation -- 2.1 Identifying the illocutionary units -- 2.2 Nucleus and ad-nuclei -- 2.1.1 Nucleus -- 2.2.2 Ad-nuclei -- 2.3 IU openers -- 2.4 Associated nuclei -- 3. Linear relations between IUs -- 3.1 Contiguous IUs -- 3.2 Embedded Illocutionary Units -- 3.3 Parentheses and bifurcations -- 3.4 Discourse units beyond IUs: The case of parallelism -- 4. The interaction between macrosyntactic and microsyntactic units -- 5. The annotation procedure -- 6. Conclusion -- 7. Annotation tools for syntax.
1. Introduction -- 2. Parsers for written and spoken French -- 2.1 Parsers for French -- 2.2 The difficulty of parsing spoken language -- 3. Segmentation and choice of a formalism -- 4. Manual annotation with Pilepilot -- 5. Unfolding-Refolding -- 6. Parsing with FRMG -- 7. Integration of FRMG into Rhapsodie's annotation process -- 8. Correction with Arborator -- 9. Agreement analysis -- 10. Post-validation correction -- 11. The distributed treebank format -- 12. Conclusion -- 8. Prosodic annotation of the Rhapsodie corpus -- 1. Introduction -- 2. Modeling and annotation of prosody: Why and how? -- 2.1 Standard models: Top-down and corpus-based strategy -- 2.2 Non-standard models: Bottom-up and corpus-driven strategy -- 3. Modeling prosody in French: The challenge of the Rhapsodie usage-based approach -- 3.1 Prominences and disfluencies: Basic features for the derivation of the prosodic structure -- 3.2 Representation of the prosodic structure: Prosodic effects of prominences and disfluencies -- 3.3 Instrumentation -- 4. Conclusion -- 9. The annotation of syllabic prominences and disfluencies -- 1. Introduction -- 2. Definitions and previous studies -- 2.1 Prominences -- 2.2 Disfluencies -- 3. Prominence and disfluency annotation in the Rhapsodie project -- 3.1 Step one: Semi-naïve annotation -- 3.2 Protocol -- 3.3 Results -- Inter-annotator reliability -- 3.4 Step two: Expert annotation -- 4. Conclusion -- 10. Segmentation into intonational periods -- 1. Introduction -- 2. General rules: Automatic segmentation into intonational periods (IPEs) -- 2.1 One speaker -- 2.2 Several speakers -- 3. Specific rules: Semi-automatic processing -- 3.1 Fillers and disfluencies -- 3.2 Global score under the threshold -- 3.3 Interactive sequences -- 3.4 Back-channels -- 4. Discussion and conclusion -- 11. Derivation of the prosodic structure -- 1. Introduction.
2. Generation of prosodic constituents inside the intonational period: Standard units -- 2.1 Metrical foot derivation (Strong_Foot, Weak_Foot): F-level rule -- 2.2 Rhythmic group derivation (Strong_Group, Weak_Group): G-level rule -- 2.3 Intonation package derivation (Included_Package): P-level-rule -- 3. Processing unattached segments: Extended units -- 3.1 Processing of disfluent segments -- 3.2 Tail labeling -- 3.3 The inclusion rule does not apply at the package level: Generation of lone and motherless packages -- 4. Discussion and conclusion -- 12. From pitch stylization to automatic tonal annotation of speech corpora -- 1. Introduction -- 2. The proposed convention for tonal annotation -- 2.1 Pitch levels -- 2.2 Pitch intervals -- 2.3 Symbols used in the annotation -- 3. Automatic labeling of prosody -- 4. The procedure for automatic tonal annotation -- 4.1 Parameter extraction -- 4.2 Segmentation into syllabic nuclei -- 4.3 Detection of pauses -- 4.4 Pitch stylization -- 4.5 Automatic detection of the pitch range of a speaker -- 4.6 Syllable-internal pitch movements -- 4.7.1 Pitch level detection based on pitch span -- 4.7.2 Pitch level detection based on local pitch change -- 4.7.3 Pitch level inferred from intra-syllabic pitch movements -- 4.7.4 Extrapolating pitch level information -- 4.7.5 Pitch levels for plateaus -- 5. The resulting tonal annotation -- 6. Discussion and conclusion -- 13. Tonal annotation -- 1. Introduction -- 2. Intonation labeling -- 2.1 Pre-processing: Speech segmentation and adaptive F0 smoothing -- 2.2 Acoustic representation of melodic contours -- 2.3 Symbolic representation of melodic contours -- 2.3.1 Frequency representation -- 2.3.2 Time representation -- 2.3.3 Formal representation -- 3. Processing interactive speech -- 3.1 Speech turns -- 3.2 Speech overlaps -- 4. Preliminary experiment -- 5. Conclusion.
14. Tools for fundamental frequency estimation in Rhapsodie -- 1. Introduction -- 2. F0 foes -- 3. Some F0 tracking methods -- 4. F0 curve cleaning -- 5. Conclusion -- 15. Exploration of the Rhapsodie corpus -- 1. Introduction -- 2. The complex data structure in Rhapsodie -- 2.1 Three independent hierarchies -- 2.2 Overlaps in prosody and syntax -- 2.3 Non-alignment of syntactic and prosodic basic units -- 3. Encoding formats and query tools -- 3.1 Tabular format -- 3.2 Trameur: A statistical query tool -- 5. Conclusion -- 16. Macrosyntax at work -- 1. Introduction -- 2. Macrosyntactic complexity -- 3. The macrosyntactic patterns identified in the corpus -- 4. Microsyntactic aspects of the most frequent macrosyntactic patterns -- 4.1 The microsyntactic nature of simple // IUs -- 4.2 The microsyntactic relation between the nucleus and the pre-nucleus in &lt -- // IUs -- 4.3 The microsyntactic relation between the nucleus and the pre-nucleus in &lt -- /&lt -- / IUs -- 4.4 The microsyntactic relation between the nucleus and the post-nucleus in &gt -- // IUs -- 4.5 The microsyntactic relation between the nucleus and the in-nucleus in ()// IUs -- 4.6 The microsyntactic internal cohesion of IUs -- 4.7 Verb-headed GUs -- 4.8 The relatively low incidence of microsyntactically incomplete IUs -- 4.9 The role of microsyntax in spoken French -- 4.10 Dependency representation and the role of microsyntax in spoken productions -- 5. Semantic aspects of the most frequent macrosyntactic patterns -- 5.1 The semantic relation between the pre-nucleus and the nucleus in &lt -- // IUs -- 5.2 The semantic relation between the pre-nuclei and the nucleus in &lt -- &lt -- // IUs -- 5.3 The semantic relation between the nucleus and the post-nucleus in &gt -- // IUs -- 5.4 The semantic relation between the in-nucleus and the nucleus in ()//IUs.
5.5 The correlation between the semantic function and the linear position of ad-nuclear constituents.
Summary: This monograph describes the development of Rhapsodie, a 33,000-word syntactic and prosodic treebank of spoken French created with the aim of modeling the interface between prosody, syntax and discourse in spoken French.
Tags from this library: No tags from this library for this title. Log in to add tags.
Star ratings
    Average rating: 0.0 (0 votes)
No physical items for this record

Intro -- Rhapsodie -- Editorial page -- Title page -- Copyright page -- Table of contents -- Acknowledgments -- Preface -- Introduction -- 1. Collecting data for the Rhapsodie treebank -- 1. Introduction -- 2. Purpose, method and context -- 2.1 Theoretical approach -- 2.2 Previous projects -- 3. Rhapsodie sampling -- 3.1 Rhapsodie corpus design: General principles -- 3.2 Gathering data: External and internal sources -- 4. Questions to be answered -- 4.1 Legal and ethical issues -- 4.2 Metadata -- 5. Conclusions -- 2. Orthographic and phonetic transcriptions of Rhapsodie recording -- 1. Introduction -- 2. From sound to orthographic transcription -- 2.1 The myth of the copyist -- 2.2. "Transcription as theory" -- 2.3 What to transcribe? -- 3. The three leading principles of orthographic transcription in Rhapsodie -- 3.1 Standard spelling -- 3.2 Punctuation and silent pauses -- 3.3 Handling of features that are specific to speech -- 3.4 Transcription conventions -- 3.5 Availability of orthographic transcription in two formats -- 4. Phonetic transcription and segmentation -- 4.1 Macro-segmentation at utterance level -- 4.2 Grapheme-to-phoneme conversion -- 4.3 Phone segmentation -- 4.4 Creation of a syllabic tier -- 5. Specific processing of multi-speaker recordings -- 6. Quality and productivity -- 7. Conclusion -- 3. Syntactic annotation of the Rhapsodie corpus -- 1. Annotating three mechanisms of cohesion and two levels of analysis -- 1.1 Microsyntax: Government and listing -- 1.1.1 Government -- 1.1.2 Listing -- 1.2 Macrosyntax -- 1.3 Government Units and Illocutionary Units -- 2. The advantages of annotating a complex structure -- 2.1 A rich syntactic treebank -- 2.2 Discourse markers, reformulations and disfluencies as part of the syntactic structure of spoken language -- 2.3 A syntactic annotation without sentences -- 3. Conclusion.

4. Microsyntactic annotation -- 1. Introduction -- 2. Morpho-syntactic Analysis -- 2.1 Word-forms and lexemes -- 2.2 Parts of speech -- 2.3 Morpho-syntactic features -- 3. Microsyntactic structure -- 3.1 Government -- 3.2 Dependency -- 3.3 Government unit -- 3.4 Choice of the head -- 4. Syntactic functions -- 5. Some complex cases -- 5.1 Extraction and qu-words -- 5.2 Negation -- 6. Phrase structure -- 7. Conclusion -- 5. The annotation of list structures -- 1. Introduction -- 2. Structure of lists -- 2.1 Layers -- 2.2 Conjuncts -- 2.3 Junctors -- 2.4 Paradigmatizing adverbs -- 2.5. List completers -- 2.6 Dependency and inherited dependency -- 2.7 Paradigmatic links -- 2.8 Junction links -- 3. Resolution of complex cases of coordination -- 3.1 Scope ambiguity -- 3.2 Gapping and non-constituent coordination -- 3.3 Embedded coordinations -- 3.4 Correlative structures -- 3.5 Junction without list -- 4. Types of lists -- 4.1 Relational coordination (para_coord) -- 4.2 Hypernymic coordination (para_hyper) -- 4.3 Intensification (para_intens) -- 4.4 Disfluency (para_disfl) -- 4.5 Reformulation (para_reform) -- 4.6 Double formulation (para_dform) -- 4.7 Negotiating formulation (para_negot) -- 5. Conclusion -- 6. Macrosyntactic annotation -- 1. Introduction -- 1.1 Macrosyntactic traditions -- 1.2 Rhapsodie's approach to macrosyntax -- 2 Rhapsodie's macrosyntactic annotation -- 2.1 Identifying the illocutionary units -- 2.2 Nucleus and ad-nuclei -- 2.1.1 Nucleus -- 2.2.2 Ad-nuclei -- 2.3 IU openers -- 2.4 Associated nuclei -- 3. Linear relations between IUs -- 3.1 Contiguous IUs -- 3.2 Embedded Illocutionary Units -- 3.3 Parentheses and bifurcations -- 3.4 Discourse units beyond IUs: The case of parallelism -- 4. The interaction between macrosyntactic and microsyntactic units -- 5. The annotation procedure -- 6. Conclusion -- 7. Annotation tools for syntax.

1. Introduction -- 2. Parsers for written and spoken French -- 2.1 Parsers for French -- 2.2 The difficulty of parsing spoken language -- 3. Segmentation and choice of a formalism -- 4. Manual annotation with Pilepilot -- 5. Unfolding-Refolding -- 6. Parsing with FRMG -- 7. Integration of FRMG into Rhapsodie's annotation process -- 8. Correction with Arborator -- 9. Agreement analysis -- 10. Post-validation correction -- 11. The distributed treebank format -- 12. Conclusion -- 8. Prosodic annotation of the Rhapsodie corpus -- 1. Introduction -- 2. Modeling and annotation of prosody: Why and how? -- 2.1 Standard models: Top-down and corpus-based strategy -- 2.2 Non-standard models: Bottom-up and corpus-driven strategy -- 3. Modeling prosody in French: The challenge of the Rhapsodie usage-based approach -- 3.1 Prominences and disfluencies: Basic features for the derivation of the prosodic structure -- 3.2 Representation of the prosodic structure: Prosodic effects of prominences and disfluencies -- 3.3 Instrumentation -- 4. Conclusion -- 9. The annotation of syllabic prominences and disfluencies -- 1. Introduction -- 2. Definitions and previous studies -- 2.1 Prominences -- 2.2 Disfluencies -- 3. Prominence and disfluency annotation in the Rhapsodie project -- 3.1 Step one: Semi-naïve annotation -- 3.2 Protocol -- 3.3 Results -- Inter-annotator reliability -- 3.4 Step two: Expert annotation -- 4. Conclusion -- 10. Segmentation into intonational periods -- 1. Introduction -- 2. General rules: Automatic segmentation into intonational periods (IPEs) -- 2.1 One speaker -- 2.2 Several speakers -- 3. Specific rules: Semi-automatic processing -- 3.1 Fillers and disfluencies -- 3.2 Global score under the threshold -- 3.3 Interactive sequences -- 3.4 Back-channels -- 4. Discussion and conclusion -- 11. Derivation of the prosodic structure -- 1. Introduction.

2. Generation of prosodic constituents inside the intonational period: Standard units -- 2.1 Metrical foot derivation (Strong_Foot, Weak_Foot): F-level rule -- 2.2 Rhythmic group derivation (Strong_Group, Weak_Group): G-level rule -- 2.3 Intonation package derivation (Included_Package): P-level-rule -- 3. Processing unattached segments: Extended units -- 3.1 Processing of disfluent segments -- 3.2 Tail labeling -- 3.3 The inclusion rule does not apply at the package level: Generation of lone and motherless packages -- 4. Discussion and conclusion -- 12. From pitch stylization to automatic tonal annotation of speech corpora -- 1. Introduction -- 2. The proposed convention for tonal annotation -- 2.1 Pitch levels -- 2.2 Pitch intervals -- 2.3 Symbols used in the annotation -- 3. Automatic labeling of prosody -- 4. The procedure for automatic tonal annotation -- 4.1 Parameter extraction -- 4.2 Segmentation into syllabic nuclei -- 4.3 Detection of pauses -- 4.4 Pitch stylization -- 4.5 Automatic detection of the pitch range of a speaker -- 4.6 Syllable-internal pitch movements -- 4.7.1 Pitch level detection based on pitch span -- 4.7.2 Pitch level detection based on local pitch change -- 4.7.3 Pitch level inferred from intra-syllabic pitch movements -- 4.7.4 Extrapolating pitch level information -- 4.7.5 Pitch levels for plateaus -- 5. The resulting tonal annotation -- 6. Discussion and conclusion -- 13. Tonal annotation -- 1. Introduction -- 2. Intonation labeling -- 2.1 Pre-processing: Speech segmentation and adaptive F0 smoothing -- 2.2 Acoustic representation of melodic contours -- 2.3 Symbolic representation of melodic contours -- 2.3.1 Frequency representation -- 2.3.2 Time representation -- 2.3.3 Formal representation -- 3. Processing interactive speech -- 3.1 Speech turns -- 3.2 Speech overlaps -- 4. Preliminary experiment -- 5. Conclusion.

14. Tools for fundamental frequency estimation in Rhapsodie -- 1. Introduction -- 2. F0 foes -- 3. Some F0 tracking methods -- 4. F0 curve cleaning -- 5. Conclusion -- 15. Exploration of the Rhapsodie corpus -- 1. Introduction -- 2. The complex data structure in Rhapsodie -- 2.1 Three independent hierarchies -- 2.2 Overlaps in prosody and syntax -- 2.3 Non-alignment of syntactic and prosodic basic units -- 3. Encoding formats and query tools -- 3.1 Tabular format -- 3.2 Trameur: A statistical query tool -- 5. Conclusion -- 16. Macrosyntax at work -- 1. Introduction -- 2. Macrosyntactic complexity -- 3. The macrosyntactic patterns identified in the corpus -- 4. Microsyntactic aspects of the most frequent macrosyntactic patterns -- 4.1 The microsyntactic nature of simple // IUs -- 4.2 The microsyntactic relation between the nucleus and the pre-nucleus in &lt -- // IUs -- 4.3 The microsyntactic relation between the nucleus and the pre-nucleus in &lt -- /&lt -- / IUs -- 4.4 The microsyntactic relation between the nucleus and the post-nucleus in &gt -- // IUs -- 4.5 The microsyntactic relation between the nucleus and the in-nucleus in ()// IUs -- 4.6 The microsyntactic internal cohesion of IUs -- 4.7 Verb-headed GUs -- 4.8 The relatively low incidence of microsyntactically incomplete IUs -- 4.9 The role of microsyntax in spoken French -- 4.10 Dependency representation and the role of microsyntax in spoken productions -- 5. Semantic aspects of the most frequent macrosyntactic patterns -- 5.1 The semantic relation between the pre-nucleus and the nucleus in &lt -- // IUs -- 5.2 The semantic relation between the pre-nuclei and the nucleus in &lt -- &lt -- // IUs -- 5.3 The semantic relation between the nucleus and the post-nucleus in &gt -- // IUs -- 5.4 The semantic relation between the in-nucleus and the nucleus in ()//IUs.

5.5 The correlation between the semantic function and the linear position of ad-nuclear constituents.

This monograph describes the development of Rhapsodie, a 33,000-word syntactic and prosodic treebank of spoken French created with the aim of modeling the interface between prosody, syntax and discourse in spoken French.

Description based on publisher supplied metadata and other sources.

Electronic reproduction. Ann Arbor, Michigan : ProQuest Ebook Central, 2024. Available via World Wide Web. Access may be limited to ProQuest Ebook Central affiliated libraries.

There are no comments on this title.

to post a comment.

© 2024 Resource Centre. All rights reserved.