lcavedon

Username: lcavedon

Participant details

Lawrence CavedonNetwork MemberLawrence Cavedon
Organisation/InstitutionRMIT University
Department/CentreDept. Computer Science and IT / Victoria Research Lab
LabNICTA
Websitehttp://www.cs.rmit.edu.au/~lcavedon
Research areaIntelligent Systems
Artificial Intelligence
Research keywords

Research Interests

My language interests focus around spoken dialogue, dialogue processes, and dialogue systems; in particular, activity-oriented dialogue for complex systems. Issues of interest include the representation of tasks, activities, goals, and intentions appropriate for effective activity-oriented conversation, as well as representation of dialogue processes themselves, using the Information-State Update paradigm. An important challenge is the use of context representation, both for robust interpretation of utterances as well as for improving interaction under error (e.g. recovering from speech mis-recognition). Of particular interest is the context of use of such systems. At Stanford, we worked on dialogue systems for the in-car environment, as well as for physical meeting spaces involving multiple participants. Different application domains throw up novel issues and constraints, such as minimising cognitive load and distraction, or managing multi-party dialogue using both spoken and visual cues.