Similarity-Based Grouping to Support Teachers on Collaborative Activities in an Exploratory Mathematical Microworld

This paper describes a computer-based tool that helps teachers group their students for collaborative activities in the context of secondary school math teaching, the challenge being to organize groups of students based on their recent work, so that their collaboration results in meaningful interactions. The complexity of the exploratory learning tasks is such that the teachers would require too long a time to create meaningful groups without the tool. This paper describes the design of the tool, the algorithms and metrics used for generating the groups, and the pedagogical context in which the tool was designed. This paper reports on the evaluation of the tool. Its recommendations are found to be equivalent to human experts, and the time required is under a few seconds for the usual classroom size. Some possibilities to extend the work to other learning domains, such as programming, are finally considered.