A Fuzzy-based Approach for Classifying Students' Emotional States in Online Collaborative Work

Emotion awareness is becoming a key aspect in collaborative work at academia, enterprises and organizations that use collaborative group work in their activity. Due to pervasiveness of ICT's, most of collaboration can be performed through communication media channels such as discussion forums, social networks, etc. The emotive state of the users while they carry out their activity such as collaborative learning at Universities or project work at enterprises and organizations influences very much their performance and can actually determine the final learning or project outcome. Therefore, monitoring the users' emotive states and using that information for providing feedback and scaffolding is crucial. To this end, automated analysis over data collected from communication channels is a useful source. In this paper, we propose an approach to process such collected data in order to classify and assess emotional states of involved users and provide them feedback accordingly to their emotive states. In order to achieve this, a fuzzy approach is used to build the emotive classification system, which is fed with data from ANEW dictionary, whose words are bound to emotional weights and these, in turn, are used to map Fuzzy sets in our proposal. The proposed fuzzy-based system has been evaluated using real data from collaborative learning courses in an academic context.