e-learning has become an important issue in the practice and research recently. A tool or system
of distance learning should provide automatic functionalities to decrease the overload of instructors
and students compared with the traditional learning style. In this paper, we present an English chat
room system in which students can discuss course contents by interacting with teachers and students.
First, the novel mechanism provides Learning_Angel Agent and Semantic Agent that act as supervisors
that constantly handle queries online. Next, the mechanism provides a QA subsystem that acts as an
assistant. The Learning_Angel Agent can detect syntax errors made by online users. The Semantic
Agent can check the semantic meaning of each sentence. Sometimes, learners make semantic level
mistakes. This implies that they do not understand the course topic or some particular issue. The
semantic agent can, thus, give some suggestions for corrections to the users and analyze the data in the
learner corpus. Moreover, when users query the system or other online users, this system will attempt
to find answers from the knowledge ontology or learner corpus that is stored in the records of previous
users’ comments. Furthermore, if enough QA pairs can be accumulated, the FAQ can act as a powerful
learning tool for learners.
Tamkang Journal of Science and Engineering 9(4) : 373-386