The aim of our research is to address the issue of distance, in online learning environments, to seek to increase the quality of communication between learner and teacher. We approach this by investigating whether intelligent system technologies can help detect disengagement and lack of motivation, and intervene strategically.
Today, there are a huge number of software solutions and intelligent approaches that have been developed for educational purposes, in order to bring the feeling of physical presence to students interacting remotely with lecturers, or to personalise their learning experience. Recently, the focus has shifted from the concept of intelligent tutors, to one of “learning companions” that have multiple roles such as that of a tutor, fellow student, mentor, and friend (Chou et al., 2003). The use of emotion in distance learning has also been explored, with a number of sophisticated systems for distance learning, such as FILTWAM (Bahreini, Nadolski, & Westera, 2014) that is able to “read” students’ emotions and correspondingly adjust the system interface in order to motivate students. However, few studies so far have paid sufficient attention to understanding how emotions impact the student-tutor relationship, and in turn the learning process. The majority of existing studies adopt a strategy of direct intervention in the student’s learning, and do not consider ways in which lecturers can be put in the picture. We seek to involve lecturers in understanding students’ emotion, and give them the opportunity to intervene if the need arises to improve students’ motivation to study.
Our research is investigating ways to capture and study the emotions of both parties, through emotion analysis, and sentiment analysis, of the textual communication between tutor and student. The sentiment/emotion analysis is also enriched by other environmental variables, such as deadlines or tasks set for the students, the classroom setting (e.g. type of class “lecture or lab”, time of the class etc.), as well as the student’s performance and behaviour thus far. This will help generate a profile that may aid the understanding on how emotions are related to performance. The expected results may help educational top management to understand the complexity of the student-lecturer relationship and its impact on emotion and environment.
The first part of the research is the technical and theoretical work underpinning the automatic detection of sentiment/emotions in the texts. We have formed a dataset, as a combination of two datasets in the literature: SemEval-2007-Task and International Survey on Emotion Detection Antecedents and Reactions (ISEAR). To this, we will add data coming from both the online learning environment that is present in our institution, and an on-ground classroom system, in use in our department, for collecting real time feedback from the class.
This work extends the research by Nabeela et al (2013) and will employ sentiment and emotion analysis tools for sentence analysis, feature extraction and classification. Case studies with on-line and on-ground students and tutors will corroborate the results.
Structure of the conference intervention: a presentation for 15 min and 5 min for discussion and questions. Understanding emotions influence students’ performance and relationship with the lecturer. So, this research falls in this theme because it uses IT to support students/staff partnership.
Altrabsheh, N, M. Gaber, and Mihaela Cocea. (2013) “SA-E: sentiment analysis for education.” 5th KES International Conference on Intelligent Decision Technologies
Chou, C. -Y., Chan, T. -W. und Lin, C. -J. (2003). Redefining the learning companion: the past, present, and future of educational agents. Computers & Education. 40, pp. 255–269.
Oluwalola, F. (2015). Effect of Emotion on Distance e-Learning — The Fear of Technology. International Journal of Social Science and Humanity, 5(11), pp. 966-970.
Yu, H., Shen, Z., Wu, Q., & Miao, C. (2014). Designing Socially Intelligent Virtual Companions. Human-Computer Interaction.