This study aims to examine studies on affective computing in e-learning. To do this, first e-learning, theoretical framework, emotions, affective computing, significance, and purpose of the study were explained. As for the method of the study, data collection tools, search details, search results, and data analysis were adverted, respectively. Following this, findings, conclusion, and suggestions were presented.

MOOCs:

A recent trend in e-learning is MOOCs. In MOOCs, free courses are delivered to a great many learners from the top-tier universities all around the world. However, MOOCs are criticized on the ground of high dropout rates, and low interaction (Morris, 2013). In helping to solve the low interaction problem, Moore’s (1989) theory of interaction seems to be appropriate.

Theoretical Framework:

Moore (1989) developed a theory about interaction in distance education. This theory classified interaction in distance education into three, which are student-content, student-teacher, and student-student interaction. This framework could help solve low interaction problem in e-learning, more specifically in MOOCs. On the other hand, another construct that should be taken into consideration to help solve the problem of high drop-out rates and low interaction in e-learning could be use of emotions.

Academic emotions:

In education, mostly the term of academic emotion is used. In e-learning, academic emotions are at the prime importance because the physical distance reduces level of transmittance of emotions between teacher and students and gives rise to negative emotions, which might lead to disengagement (Qin et al., 2010). The most popular academic emotion, on the other hand, is anxiety (Pekrun, Goetz, Titz & Perry, 2002). However, boredom and enthusiasm are also considered important emotions in e-learning.

Affective Computing:

The application of academic emotions to e-learning is said to be affective computing. In affective computing, the main purpose is to detect students’ or even teacher’s emotions during an e-learning course through eye-movements, head movements, voice, facial gestures using eye-tracking devices, cameras, microphones and advanced software. Affective computing is highly related to adaptive learning systems because e-learning courses can be customized for students utilizing data collected through affective computing.

Justification and Purpose:

Emotions play a key role in e-learning process. Recent studies have shown the effect of emotions on academic achievement (D'Errico, Paciello & Cerniglia, 2016; Krithika & Priya, 2016; Qin, Zheng, Li, & Zhang, 2010). In these studies, it was pointed out that positive emotions increase learning, enthusiasm, and engagement.

Because of the reasons mentioned above, emotion detecting systems are intensely studied in the literature. Through using these systems, detecting and evaluating emotions could help improve e-learning experience of learners by disseminating content, and giving real time and delayed feedback tailored for learners (Krithika & Priya, 2016; Ray & Chakrabarti, 2012). In short, detecting emotions in e-learning could help improve effectiveness of e-learning through adaptive learning systems (Caballé et al., 2014).

As there are a very large number of studies on affective computing in e-learning, there is a need for a quantitative content analysis on affective computing studies in e-learning. Therefore, a quantitative content analysis regarding aforementioned topic was carried out to lead the way for both implementers and researchers. To that end, following six research questions were answered.

1) What are the years in which emotion-related studies in e-learning were published?

2) Which publication types are preferred in publishing emotion-related studies in e-learning?

3) Which academic units conduct studies on emotion-related studies in e-learning?

4) Which countries are leading the emotion-related studies in e-learning?

5) What are the main purposes of studies conducted on emotion-related issues in e-learning?

6) What are the correct detection ratios of emotions in emotion-related studies in e-learning?

A quantitative content analysis was carried out in the present study. Quantitative content analysis is used to classify data stemming from pre-determined concepts and themes for the sake of clarity (Fraenkel & Wallen, 2000). Data Collection Instruments: A data collection tool developed by the researchers and named as “publication examination form” was utilized. In forming this tool, a literature review was conducted. After this literature review, critical information that needs to be collected was determined. This form contains six critical information about publications, which are 1) publication year, 2) publication type, 3) researchers’ academic unit, 4) country of publication, 5) main purpose of publication and, 6) correct detecting ratios of emotions. Search Details: Web of science publication database, which is believed to be the most comprehensive source for indexed reference searching, was preferred to conduct the search for this study. Proceedings and ESCI (Emerging Science Citation Index) were included in the search. Search terms of emotion, feeling, affect, mood, and sentiment with ”OR” search operator between them were used. These search terms were also used together with search term of e-learning connected with the search operator of “AND”. This search, conducted in January 08, 2017, was limited with the title section of the studies. It was also limited with the years of 1945 through 2017. Search Results and Data Analysis: Obtained publications subsequent to search were first analyzed in regard to relevance. Irrelevant publications were discarded from the study. Following this, they were summarized in 50-75 words. 72 publications were obtained subsequent to search. After the examination of the studies in terms of relevance, which meant the elimination of 37 publications, the study was carried out with 35 publications. As for analysis part, frequencies and percentiles were used.

It was seen that correct detection ratios of emotions have steadily soared to nearly 100% in recent publications. As studies on affective computing in e-learning started around 2006, there has been a remarkable technological advance helping this ratio soar over the last decade. It was found out that studies on affective computing in e-learning are mostly carried out by technology related academic units. This is because, affective computing requires handling highly technical software such as audio and visual processing. It was evident that the main purpose of the majority of publications was to propose a system detecting learners’ emotions in e-learning. This is because the prime purpose of current e-learning initiatives is to customize e-learning experience. E-learning experience could be easily customized with adaptive learning environments through detected emotions of learners. Suggestions: In the present study, the search was conducted in only Web of Science academic database. Further studies could do search in other academic databases such as Science Direct, SCOPUS, and ERIC. Since in this study mostly proceedings were examined, further studies could focus only on articles or dissertations. The search could be done in the topic field unlike this study. In further content analyses, the keywords of online learning, distance learning, and web-based learning could be used in addition to e-learning. As for keywords regarding emotions, more specific emotions such as happiness, satisfaction, trust, boredom, anxiety, and interest could be preferred. Given the growing popularity of MOOCs and affective computing over the last decade, studies connecting these two on the ground of adaptive learning environments could be conducted. Customization arising from the applications of adaptive learning systems through the output of affective computing such as affective feedback could help solve the problems associated with MOOCs like high drop-out rates and low interaction.

Caballé, S., Barolli, L., Feidakis, M., Matsuo, K., Xhafa, F., Daradoumis, T., & Oda, T. (2014). A study of using SmartBox to embed emotion awareness through stimulation into e-learning environments. Proceeding of the 2014 International Conference on Intelligent Networking and Collaborative Systems (INCoS), (pp. 469–474). Salerno, Italy: IEEE. D'Errico, F., Paciello, M., & Cerniglia, L. (2016). When emotions enhance students’ engagement in e-learning processes. Journal of e-Learning and Knowledge Society, 12(4), 9-23. Fraenkel, J. R., & Wallen, N. (2000). How to design and evaluate research in education (4th ed.). NY: McGraw-Hill. Krithika, L. B., & Priya, G.G.L. (2016). Student emotion recognition system (SERS) for e-learning improvement based on learner concentration metric. In S.A. Ibrahim, S. Mohammad, & S.A. Khadar (Eds.), Proceedia Computer Science: Vol. 85. International Conference on Computational Modelling and Security (CMS 2016) (pp. 767-776). Bengaluru, India. doi: 10.1016/j.procs.2016.05.264. Moore, M. G. (1989). Editorial: Three types of interaction. American Journal of Distance Education, 3(2), 1–6. Morris, L. V. (2013). MOOCs, emerging technologies, and quality. Innovative Higher Education, 38(4), 251–252. doi: 10.1007/s10755-013-9263-2. Pekrun, R., Goetz, T., Titz, W., & Perry, R. P. (2002). Academic emotions in students' self-regulated learning and achievement: A program of qualitative and quantitative research. Educational psychologist, 37(2), 91-106. Qin, J., Zheng, Q., & Li, H. (2014). A study of learner-oriented negative emotion compensation in e-learning. Educational Technology & Society, 17(4), 420-431. Qin, J., Zheng, Q., Li, H., & Zhang, H. (2010). An emotion regulation model in an e-learning environment. Proceeding of the 9th International Conference on Web-Based Learning (pp. 240-249). Shanghai: China: Springer Berlin Heidelberg. Ray, A., & Chakrabarti, A. (2012). Design and implementation of affective e-learning strategy based on facial emotion recognition. Proceedings of the 1st International Conference on Information Systems Design and Intelligent Applications (INDIA 2012) (pp. 613–622). Visakhapatnam, India: Springer Berlin Heidelberg.