Classifying speech acts using verbal response modes

Andrew Lampert, Robert Dale, Cecile Paris

Research output: Chapter in Book/Report/Conference proceedingConference proceeding contributionpeer-review


The driving vision for our work is to provide intelligent, automated assistance to users in understanding the status of their email conversations. Our approach is to create tools that enable the detection and connection of speech acts across email messages. We thus require a mechanism for tagging email utterances with some indication of their dialogic function. However, existing dialog act taxonomies as used in computational linguistics tend to be too task- or application-specific for the wide range of acts we find represented in email conversation. The Verbal Response Modes (VRM) taxonomy of speech acts, widely applied for discourse analysis in linguistics and psychology, is distinguished from other speech act taxonomies by its construction from crosscutting principles of classification, which ensure universal applicability across any domain of discourse. The taxonomy categorises on two dimensions, characterised as literal meaning and pragmatic meaning. In this paper, we describe a statistical classifier that automatically identifies the literal meaning category of utterances using the VRM classification. We achieve an accuracy of 60.8% using linguistic features derived from VRM’s human annotation guidelines. Accuracy is improved to 79.8% using additional features.
Original languageEnglish
Title of host publicationProceedings of the 2006 Australasian Language Technology Workshop
EditorsLawrence Cavedon, Ingrid Zukerman
Place of PublicationSydney Australia
Number of pages8
ISBN (Print)1741081467
Publication statusPublished - 2006
EventAustralasian Language Technology Workshop (ALTW) 2006 -
Duration: 30 Nov 20061 Dec 2006


WorkshopAustralasian Language Technology Workshop (ALTW) 2006


  • Verbal Response Modes (VRM)

Fingerprint Dive into the research topics of 'Classifying speech acts using verbal response modes'. Together they form a unique fingerprint.

Cite this