Making Virtual Learning Environment more intelligent: the problem of software agent’s mental state recognition

Abstract

Intelligent decision making process, which is performed according to a learner in the Virtual Learning Environment (VLE), leads to the problem of solving several rather complex tasks. Two of them are most of interest. First, we need to train a software agent to recognize its mental state; and second, we want agent to apply optimal strategy to teach learners when it is in some mental state. The second issue we have discussed in [2]. We propose the agent’s ability to recognise its mental state could be based on the classification result to a particular learner. Classification used to be done should be based on the discovered groups in the database data and classification rules that prescribe given learner to the one of the available clusters. Conceptual clustering seems to be the suitable technique able to provide a solution for the problem raised.

Keywords

• agent’s state
• clustering
• conceptual clustering
• machine learning
• virtual learning environment