Including malicious agents into a collaborative learning environment

Abstract

In this paper we introduce a collaborative environment for the development of medium-size programming projects. Our system provides the usual facilities for communication among members of the group as well as a friendly programming environment for the functional programming language Haskell. A relevant feature of our learning environment is that some of the students may be, in fact, virtual students. It is worth to point out that these agents will not always behave as helpers. On the contrary, it can happen that they produce, on purpose, wrong programs. By doing so, we pretend that students get the abilities to detect mistakes not only in their own code, but also in the code generated by other team-mates.