Teachable Agent

• Published in: arXiv.org
• Main Author: Borjigin, Ailiya
• Format: Paper
• Language: English
• Published: Ithaca Cornell University Library, arXiv.org 09.02.2015
• Subjects: Authoring; Education; Knowledge base; Learning; Learning by teaching; Pedagogy; Students; Teaching
• ISSN: 2331-8422

Teachable Agent (TA) is a special type of pedagogical agent which instantiates the educational theory of Learning by Teaching. Soon after its emergence, research of TA becomes an active field, as it can solve the over scaffolded problem in traditional pedagogical systems, and encourage students to take the responsibility of learning. Apart from the benefits, existing TA design also has limitations. One is the lack of enough proactive interactions with students during the learning process, and the other is the lack of believability to arouse students empathy so as to offer students an immersive learning experience. To solve these two problems, we propose a new type of TA, Affective Teachable Agent, and use a goal oriented approach to design and implement the agent system allowing agents to proactively interact with students with affective expressions. The ATA model begins with the analysis of pedagogical requirements and teaching goals, using Learning by Teaching theory to design interventions which can authentically promote the learning behaviors of students. Two crucial capabilities of ATA are highlighted Teachability, to learn new knowledge and apply the knowledge to certain tasks, and Affectivability, to establish good relationship with students and encourage them to teach well. Through executing a hierarchy of goals, the proposed TA can interact with students by pursuing its own agenda. When a student teaches the agent, the agent is performed as a naive learning companion, and when an educator teaches the agent during the design and maintenance time, the agent can perform as an authoring tool. To facilitate the involvement of educators into the game design, we develop an authoring tool for proposed ATA system, which can encapsulate the technical details and provide educational experts a natural way to convey domain knowledge to agent knowledge base.