Intelligent support for group work in collaborative learning environments

摘要: The delivery of intelligent support for group work is a complex issue in collaborative learning environments. This particularly pertains to the construction of effective groups and assessment collaboration problems. because the composition can be affected by several variables, various methods are desirable ascertaining existence different problems. Literature has shown that current collaborative environments provide limited or no teachers cope with these tasks. Considering this the increasing use online collaboration, research aims explore solutions for improving in learning environments, thus simplify how manage work. In thesis, three aspects were investigated achieve goal. first aspect emphasises on proposing novel approach formation based on students‘ styles. novelty importance the provision an automatic grouping method tailor individual students‘ characteristics fit well into existing environments. The evaluation activities comprise development add-on tool an undergraduate student experiment, which indicate feasibility strength the proposed — being capable forming diverse tend perform more effectively efficiently than similar conducting group discussion tasks. The second focus relates identification major group problems their causes. A nationwide survey was conducted that reveals perspective issue, literature fails to adequately address. Based findings from survey, XML-based representation created provides unique linkages between causes identified. Finally, then shifted proposal for diagnosing major identified. originality and significance lies provision methods identified, on student interaction data result examined. evaluation procedure focused supporting several experiments test dataset. results show the feasibility effectiveness are sustained, great extent, diagnostic methods addressed. Besides main proposals, explored multi-agent architecture unify all components derived intelligently managing online collaborative learning, suggests overarching framework providing context other parts thesis.