An ubiquitous teaching assistant using knowledge retrieval and adaptive learning techniques

Learning environments with rich interactivity, friendly communication interface, as well agents for instruction assistance have become the goal many distance education applications seek to achieve. In this paper, we have designed and implemented a virtual teaching assistant agent that possesses both domain-specific course-related knowledge, as well as the intelligence to handle daily conversation. Constructed using standardized knowledge base representation, the virtual teaching assistant agent's knowledge is thus replaceable, making it reusable in multiple courses. Carried in both Web as well as modern instance message interfaces, the agent supports both keyword and sentence based query, as well as multi-lingual functionality. In addition, it owns the ability to learn from other agents or knowledge sources, thus enriching its intelligence. The experimental results showed that a pedagogical agent with adaptive learning and communication capabilities indeed has a significant improvement to the quality of modern distance education.