Elamur Research⌕
☰
home / papers / 2009 / Towards a Multidimensional Semantics of Discourse Markers in Spoken Dialogue

Towards a Multidimensional Semantics of Discourse Markers in Spoken Dialogue

Summary

This paper argues that discourse markers in spoken dialogue (words such as "and", "but", "so") often perform several communicative jobs at once, and that this multifunctionality is best captured with a multidimensional semantic model. The authors use Dynamic Interpretation Theory (DIT), which treats a dialogue utterance as a bundle of context-update operations spread over separate dimensions, such as Task, Auto-/Allo-feedback, Turn Management, Time Management, and Dialogue Structuring. In this view, a marker can act as a dialogue act in its own right, or it can help determine the function of a general-purpose or dimension-specific act. The study uses AMI meeting recordings and examines "and" in detail. It reports statistical analyses of prosodic and lexical features and runs machine-learning experiments with the RIPPER rule inducer, evaluated by 10-fold cross-validation. Discourse markers can be automatically separated from non-marker uses, and their functions in most dimensions can be predicted above the majority-class baseline. The authors conclude that markers carry functions in nearly every DIT dimension and help delimit functional segments. This is relevant for dialogue segmentation and computational dialogue modelling.

Key points

  • Framework: DIT distinguishes general-purpose functions, such as Inform, Suggestion, or Elaborate, from dimension-specific ones, such as Turn Keep, Stalling, or topic shift. Earlier work cited in the paper found that a functional segment has on average at least 1.3 functions.
  • DM vs. non-DM classification: The majority-class baseline for distinguishing discourse-marker uses of "and" was 54.5%.
    • Marker uses last roughly 289–327 ms, compared with 173–217 ms for non-marker uses.
    • Marker uses are preceded by pauses of about 59–228 ms.
    • Marker uses have a higher mean pitch, by more than 12 Hz.
  • Turn keeping: In 57% of cases, "and" signals that the speaker intends to keep the turn.
  • Functional classification: RIPPER beat the baseline in most dimensions. The exceptions were Discourse Structuring and Partner Communication Management, which had too few training instances.
    • Lexical cues (bag-of-words, collocations) mattered most for Task functions.
    • Prosody, together with speaker information, dominated in the other dimensions. For example, Stallings averaged about 585 ms in duration with a 365 ms initial pause.
  • Coverage and future work: Markers appeared in all DIT dimensions except possibly Social Obligations Management. Planned next steps are larger annotated data and multimodal cues such as gestures from the AMI video recordings.

Our summary of the paper, in our own words — not its text. The work belongs to its authors; read it at the links above.

Fields: Linguistics & languages

Topics: Discourse analysis

Cited by

5 papers of our library cite it (of the 140673 whose bibliographies we read); 2 of them have a page here, 3 not yet.

Cite

The paper itself, not this page. BibTeX from the library's record of the work.

@inproceedings{petukhova2009multidimensional,
  title = {{Towards a Multidimensional Semantics of Discourse Markers in Spoken Dialogue}},
  author = {Petukhova, Volha and Bunt, Harry},
  booktitle = {Proceedings of the Eighth International Conference on Computational Semantics},
  year = {2009},
  doi = {10.3115/1693756.1693773},
  url = {https://doi.org/10.3115/1693756.1693773}
}

References

16 references, 1 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Allwood, J. (1992). On dialogue cohesion. Gothenburg Papers in Theoretical Linguistics 65. Gothenburg University, Department of Linguistics. find
  2. Allwood, J. (2000). An activity-based approach to pragmatics. H. Bunt and W. Black (eds.) Abduction, Belief and Context in Dialogue, Amsterdam: Benjamin, pp.47-81. ↗
  3. Bunt, H. (2000). Dynamic Interpretation and Dialogue Theory. M.M. Taylor, D.G. Bouwhuis and F. Neel (eds.) The Structure of Multimodal Dialogue, Vol 2., Amsterdam: John Benjamins, pp. 139–166. find
  4. Bunt, H. (2006). Dimensions in Dialogue Act Annotation. Proceedings LREC 2006, Genova. find
  5. Bunt, H. (2007). Multifunctionality and Multidimensional Dialogue Act Annotation. E. Ahlsen et al. (ed.) Communication - Action - Meaning. Gothenburg, pp. 237 – 259. find
  6. Cohen, R. (1984). A computational theory of the function of clue words in argument understanding. Coling-ACL 1984, Standford, pp. 251–258. find
  7. Cohen, W.W. (1995). Fast effective rule induction. In Proceedings of the 12th International Conference on Machine Learning (ICML’95), pp. 115–123. ↗
  8. Geertzen, J. Petukhova, V. and Harry Bunt. (2007). A Multidimensional Approach to Utterance Segmentation and Dialogue Act Classification. Proceedings of the 8th SIGdial Workshop on Discourse and Dialogue, Antwerp, pp. 140–149. find our card
  9. Heeman, P.A. and Allen, J.F. (1999). Speech repairs, intonational phrases and discourse markers: Modelling speakers utterances in spoken dialogue. Computational Linguistics, 12(3): 1–45. find
  10. Hirschberg, J. and Litman, D. (1993). Empirical studies on the disambiguqtion of cue phrases. Computational Linguistics, 25(4): 501–530. find
  11. Hovy, E.H. (1995). The multifunctionality of discourse markers. Proceedings of the Workshop on Discourse Markers, Egmond-aan-Zee, The Nertherlands. find
  12. Mann, W. and Thompson, S. (1988). Rhetorical structure theory: toward a functional theory of text organisation. The MIT Press, Cambridge, MA. find
  13. Morante, R. (2007). Computing Meaning in Interaction. PhD Thesis, Tilburg University. find
  14. Popescu-Belis, A. and Zufferey, S. (2006). Automatic identification of discourse markers in multiparty dialogues. Working paper 65, ISSCO, University of Geneva. find
  15. Schiffrin, D. (1987). Discourse Markers. Cambridge: University Press. find
  16. Sporleder, C. and Lascarides, A. (2008). Using Automatically Labelled Examples to Classify Rhetorical Relations: An Assessment. Natural Language Engineering, 14.03: 369–416. find

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/petukhova-2009-multidimensional-semantics-discourse-markers-spoken/