Cognitively ergonomic analogical explanations in conversations about everyday tasks

Researcher

Lina Mavrina

Publications

• Lina Mavrina, Stefan Kopp (2023): Predicting grounding state for adaptive explanation generation in analogical problem-solving Full Publication

Collaboration

Judith Sieker

Research Theme

R1 Human agency

We are building an assistive system with a goal to support its users in the accomplishment of everyday problem-solving tasks in the long-term perspective, such as using appliances at home. The system interacts with the users by means of spoken dialogue and employs adaptive analogical explanations to build up on the existing knowledge of the user and help them transfer it to a new situation. By using analogies as a natural means of explanation, we aim to make the system as cognitively ergonomic as possible. We are also researching the means of evaluating such a system and measuring the success of its use under field conditions.

Both domain-knowledge-related and social-interaction-related aspects of explanation planning play an important role in finding an appropriate analogical explanation. The interplay between these aspects is best illustrated in the structure of system’s conversational memory which is a graph-based representation integrating domain knowledge with dialogue state and history information. This representation and the mechanisms allowing the system to reason over it during explanation planning are the focus of our research in SAIL.