Automating the formation of sporadic groups and the selection of topics for online language learning driven by personality traits and interest profiles

© 2016 IEEE. Sporadic Groups for Online Language Learning (SGOLL) are becoming an increasingly popular service. Nowadays there are many portals whose business is to provide a meeting point for teachers and students, as well as the technological platform to support the online conversation spaces. The students' Quality of Experience (QoE) is of utmost importance for the companies running those portals, because -as it happens in many other areas of the Web 2.0- potential new users pay much attention to the comments and ratings provided by others in the past. Accordingly, the companies implement mechanisms for the students to rate the teachers, in order to promote the ones who get the best evaluations. In this doctoral research, we address the question of whether they could take additional (proactive) measures in quest for greater satisfaction, trying to maximise the chances (i) that the people who gather together in a conversation session will get on well together, and (ii) that they will have interesting topics to talk about. We are developing solutions to aid in the planning of the sessions, leaning on data mining and machine learning technologies.