Dialogue and information flow

This theme focuses on computational models of dialogue and information flow. It includes research on multimodal reference, inference and questions in dialogue, dialogue script generation, as well as foundational research into what counts as speaking a natural language and how to model this computationally.

Key publications

• Are conversational large language models speakers?
• Discourse annotation - Towards a dialogue system for pair programming
• Challenges with Learning to Program and Problem Solve: An Analysis of Student Online Discussions
• Dialogue with computers: dialogue games in action
• Towards a computational account of inferentialist meaning
• Dialogue structure and logical expressivism
• Data-oriented monologue-to-dialogue generation
• Salience in the generation of multimodal referring acts
• Presenting Arguments as Fictive Dialogue

Projects

• Multimodal Dialogue System for Pair Programming (2022 - 2026)
• COherent Dialogue Automatically generated from text (CODA) (2009 - 2011)

Resources

• Inferentialism: Inferentialist Semantics and Pragmatics