Analysis of collaboration in CSCL chat using rhetorical schemas

Chat is considered one of the most important and popular tools that are widely used in collaborative processes. Although various applications are used in Computer-Supported Collaborative Learning (CSCL), only few applications can help teachers in assessing the chats produced by learners. The fact that chats are rarely used in the official educational context may be explained by the difficulty of their grading. To find a solution for this problem, this paper presents a model that is assumed to enable teachers in assessing chats semi-automatically. It also includes the results of the analysis of three chats in the form of statistics and graphics, which can help teachers in easily assessing. Therefore, this model served the analysis of the contents of the chats, giving an example of how to assess the collaborative and individual participations. These conversations were conducted between a group of learners who were discussing predefined subjects in educational sessions. Moreover, the proposed model is based on assessing individual participations through identifying the most important threads discussed by each participant in the chat, relying on Mikhail Bakhtin's dialogism theory and Stefan Trausan-Matu's polyphonic model. The model used an effective way to find associative relationships among the important threads discussed by students and the rhetorical structures, considering linking cue phrases (`but', `therefore', `and', `or', ...). Starting from it, collaboration is assessed by studying the relations resulted from linking the important threads using cue phrases.