Affective Computing and Motivation in Educational Contexts: Data Pre-processing and Ensemble Learning

Abstract

Affective computing involves the research area focusing on the implementation of systems and mechanisms that are capable of identifying, understanding and elaborating human affections. It is a field that spans computer science, psychology and cognitive science. Nowadays, affective computing is considered to play a vital role in the field of e-learning since knowledge acquisition can be greatly affected by the changing emotional states of learners. The machines should understand the emotional states of students and adapt their behavior to them, thus providing a tailored response to these emotions. Affective computing systems identify the user’s emotional state and react accordingly. In view of the above, this chapter includes a short presentation of the concepts of affective computing tailored to social networking-based learning and learners’ affective states. Special mention is made of students’ frustration as an emotional state that can either influence the students’ learning rates or dropout rates and motivation strategies to overcome problems emerging from negative emotions. This chapter also presents well-known motivational theories as well as pre-processing techniques and ensemble classifiers for sentiment analysis through social networks. Motivation theories concern the support of students to achieve a goal or a certain performance level, leading to goal-directed behaviors. Pre-processing techniques deal with the necessary information to preprocess the reviews in order to find sentiment and make analysis whether it is positive or negative. Finally, Sentiment analysis refers to the use of expert methods (such as natural language processing, text analysis, computational linguistics) to systematically identify, extract, quantify, and study affective states and subjective information.