Abstract—This paper presents a study about which kind of interaction leads to better learning within a collaborative learning task, in terms of the effects of a speech act on both the actor and their collaborative partner, in a cooperative or competitive behavior. To perform this study, we have developed an analysis model that investigates features in dialogue interactions based on student’s differences in understanding, as shown by their conversation protocol, and an investigation of the cognitive conflict based on Piagetian theories. We have analyzed data from a study conducted in a computer supported collaborative learning environment (CSCL) in the domain of fractions. The long-term goal of this model is to better understand collaborative learning and to support pedagogical agents in intervening in dialogue in order to keep collaboration productive.

Index Terms—CSCL, Education, NLP, Assessment, Pedagogical Agents.

David Prata and Patrick Letouze are with the Federal University of Tocantins, Palmas, CO 77.001-019 BRAZIL (e-mail: ddnprata@ uft.edu.br,letouze@ uft.edu.br).
Evandro Costa is with the Federal University of Alagoas, Maceio, CO57.072-900 BRAZIL (email:ebcosta@gmail.com).

Cite: David Prata, Patrick Letouze, Evandro Costa, Monica Prata, and George Brito, "Dialogue Analysis in Collaborative Learning," International Journal of e-Education, e-Business, e-Management and e-Learning vol. 2, no. 5, pp. 365-372, 2012.