Analysis of interactions in a collaborative learning session of Instant Messaging

This paper analyzes, from the polyphonic perspective used in the PolyCAFe system, the interactions among a group of high school students who, in parallel with a classroom style of interaction, use the instant messaging system Yahoo Messenger to develop a sorting algorithm. This type of analysis aims at finding the main topics in a chat conversation, the evaluation of utterances and their implicit links. Moreover, are investigated the ways in which the analysis can be automated using natural language processing tools. In addition to the previous research on using the polyphonic approach, the paper introduces the idea that pauses are important and should also be considered in the analysis. In collaborative hybrid (both face-to-face and online at a distance) learning educational environments one of the most used type of tools are the chat conversations. We can therefore consider Computer Supported Collaborative Learning (CSCL) as an important branch of hybrid learning.