Smart and Interactive e-Learning System Based on Smartive
Nobukazu Yoshioka, Shinichi Honiden

E-Learning has been important role for teaching English conversation. Traditional e-Learning systems, however, have problems from the viewpoints of system developers, teachers and learners. From the developers' point of view, it is difficult to develop systems which dynamically change contents to adapt to the ability and situation of learners. From the teachers' point of view, it is difficult to ensure consistency with policies or guidelines of each teacher. Additionally, it is difficult to teach English conversation with group learning because of difference of conversation ability. From the learners' point of view, e-Learning tends to be tired easily because of lack of the interaction with people.
We have proposed a new e-Learning system overcoming such problems using Smartive. Smartive is an agent middleware which we have been developing for content delivery, which uses agent technologies such as self-control and collaboration policies and negotiation among agents. The e-Learning system has four kinds of functions. First, we can adapt to different learners levels. In other words, contents can be automatically changed with respect to learner's levels. Secondly, it can adapt to learn with a group to avoid losing the interest. Thirdly, it can decide conversation partners based on both teachers and learners policies. Fourthly, advice of teachers can be inserted automatically between conversations with respect to the progress of conversation and learners ability.
We have evaluated the ability through experiments with junior high school classes. In consequence, students felt appropriate difficulty with the conversation and learned without losing the interest. In addition, teachers evaluated the usefulness of automatic insert of advice based on their guidelines. After all, the system allowed the conversation with learners with difference levels, which the conversation partners was selected based on both teachers and learners wishes.

Nobukazu Yoshioka is a researcher at the National Institute of Informatics, Japan.  Dr. Nobukazu Yoshioka received his B.E degree in Electronic and Information Engineering from Toyama University in 1993.  He received his M.E. and Ph.D. degrees in School of Information Science from Japan Advanced Institute of Science and Technology in 1995 and 1998, respectively.  From 1998 to 2002, he was with Toshiba Corporation, Japan. From 2002 to 2004 he was a researcher, and since August 2004, he has been an associate professor, in National Institute of Informatics, Japan.  His research interests include agent technology, object-oriented methodology, software engineering, and software evolution.  He is a member of the Information Processing Society of Japan (IPSJ), and Japan Society for Software Science and Technology.