Case-based reasoning approach for utilisation of past remarks as advice for collaborative learning

The collaborative learning remarks that were triggers of the past active discussion, which is called discussion promotion remark, can also be effective in the current discussion situation whose speech patterns are similar. The objective of this research is to construct the advising system for the collaborative learning which utilises discussion promotion remarks of the past collaborative learning as advice to solve current inappropriate situation. Discussion consists of various characteristics, so it is difficult to define appropriate speech pattern for each type of discussion promotion remark. This research introduces case-based reasoning approach to extract past discussion promotion remark which can solve current inappropriate situation. This paper describes two of the steps to accomplish the case-based reasoning system. First, several parameters that characterise the discussion situation are introduced and attached to past discussion promotion remarks. Second, discussion promotion remark database is constructed as a decision tree based on the attached parameters.