A Configurable Conversational Agent to Trigger Students’ Productive Dialogue: A Pilot Study in the CALL Domain

Research areas:

• Uncategorized

Year:

2013

Type of Publication:

Article

Keywords:

conversational agents, collaborative learning, productive dialogue

Authors:

• Tegos, Stergios
• Demetriadis, Stavros
• Tsiatsos, Thrasyvoulos

Pages:

581-584

Abstract:

MentorChat is a CSCL system that utilizes a conversational pedagogical agent to engage students in fruitful collaboration. This conversational agent attempts to promote students’ productive dialogue by providing unsolicited prompts aimed to elicit explicit reasoning. Using MentorChat a teacher is able to design and deploy the phases and the tasks of a collaborative activity along with the support that the agent provides. In this paper, we present (a) the theoretical background Men-torChat is based on, (b) the main technical features of the system and (c) some encouraging preliminary results