Automatic Generation of Dialogue Acts: Understanding Dialogue through Computational Modelling

Topic Description

The purpose of this project is to gain a better understanding of human-human dialogue through formal/computational modelling (Piwek, 2017; 2011; 2006). The project will focus on one out of the following three phenomena in dialogue:

1. multimodal dialogue acts (involving gesture and speech) (Piwek 2009; Piwek, Beun & Cremers, 2008),
2. generation of questions (Piwek & Boyer, 2012; Kuyten et al., 2012; Rus et al. 2012; Wyse & Piwek, 2009), or
3. adversarial dialogue moves (and emotion) (Pluss and Piwek, 2016).

The main aim of the project will be to construct a computational model of one of these phenomena. This will involve building algorithms for automatically generating dialogue acts that exhibit the relevant phenomenon (for example, a model for automatically generating questions or, for instance, for generating multimodal dialogue acts).

The research will involve developing a pair of agents that have the ability to converse with each other, with the resulting behaviour mirroring, in relevant respects, the human behaviours as we find them in naturally-occurring dialogues.

Skills Required:

A good degree (2:1 or above) in Computing (or equivalent), with experience in artificial intelligence, computational linguistics or natural language processing. A background in linguistics would also be suitable, if accompanied by relevant computer skills.

Background Reading:

Domingo, Cecilia; Piwek, Paul; Stoyanchev, Svetlana and Wermelinger, Michel (2023). Discourse annotation - Towards a dialogue system for pair programming. TAL Journal, 63(3) pp. 11–35. 

Kuyten, Pascal; Bickmore, Timothy; Stoyanchev, Svetlana; Piwek, Paul; Prendinger, Helmut and Ishizuka, Mitsuru (2012). Fully automated generation of question-answer pairs for scripted virtual instruction. In: 12th International Conference on Intelligent Virtual Agents, 12 - 14 September 2012, Santa Cruz, CA, USA. https://oro.open.ac.uk/34669/

Piwek, Paul (2006). Perspectives on dialogue: Introduction to this special issue. Research on Language and Computation, 4(2-3) pp. 143–152. https://oro.open.ac.uk/3187/

Piwek, Paul; Beun, Robbert-Jan and Cremers, Anita (2008). 'Proximal' and 'distal' in language and cognition: Evidence from deictic demonstratives in Dutch. Journal of Pragmatics, 40(4) pp. 694-718. https://oro.open.ac.uk/12183/

Piwek, Paul (2009). Salience and pointing in multimodal reference. In: Proceedings of Production of Referring Expressions: Bridging the gap between computational and empirical approaches to generating reference (PRE-CogSci 2009), 29 July 2009, Amsterdam. https://oro.open.ac.uk/19167/

Piwek, Paul (2011). Three principles of information flow: conversation as a dialogue game. In: Ramage, Magnus and Chapman, David eds. Perspectives on Information. Routledge Studies in Library and Information Science (9). New York, USA: Routledge, pp. 106-120. 

Piwek, Paul and Boyer, Kristy Elizabeth (2012). Varieties of Question Generation: introduction to this special issue. Dialogue & Discourse, 3(2), pp. 1-9. https://oro.open.ac.uk/33237/

Piwek, Paul (2017). Dialogue with computers: dialogue games in action. In: Mildorf, Jarmila and Thomas, Bronwen eds. Dialogue across Media. Dialogue studies (28). Amsterdam: John Benjamins, pp. 179–202.

Plüss, Brian and Piwek, Paul (2016). Measuring Non-cooperation in Dialogue. In: Proceedings of the 26th International Conference on Computational Linguistics (COLING 2016), 11-16 Dec 2016, Osaka, Japan.

Rus, Vasile; Wyse, Brendan; Piwek, Paul; Lintean, Mihai; Stoyanchev, Svetlana and Moldovan, Cristian (2012). A detailed account of the First Question Generation Shared Task Evaluation challenge. Dialogue & Discourse, 3(2) pp. 177–204. https://oro.open.ac.uk/33235/

Wyse, Brendan and Piwek, Paul (2009). Generating questions from OpenLearn study units. In: AIED 2009 Workshop Proceedings Volume 1: The 2nd Workshop on Question Generation, 6-9 July 2009, Brighton, UK. https://oro.open.ac.uk/32435/