Analysis of Emotion Recognition Methods: A Systematic Mapping of the Literature

LAURA ABALLAY; CESAR COLLAZO; SILVANA ACIAR; ALEX TORRES

Telematics and Computing

Springer Cham

Año: 2024; p. 298 - 313

Research focuses on analyzing the existing literature on emotion detection methods and their applicability in Virtual Learning Environments (VLEs). The objective is to provide a comprehensive and categorized overview of the literature on emotion detection methods, identifying the technologies and methodologies used, the most commonly detected emotions, and the suitability of these methods for application in VLEs. The method included an automated search of articles from January 2016 to April 2023, both in English and Spanish. The databases used were ScienceDirect, IEEE Explore, and ACM Digital Library. The search terms focused on emotion detection in educational settings. The results showed a variety of emotion detection methods, such as facial analysis, voice analysis, physiological sensors and text analysis. The most detected emotions were happiness, sadness, boredom, and amusement. Some methods proved to be particularly suitable for implementation in EVAs. In the conclusions, the increasing diversity of methods and technologies for emotion detection is highlighted, underlining areas that require further research. It is recommended that less invasive and more accurate technologies be developed and that ethical and privacy issues associated with emotion detection in educational settings be addressed. In addition, it is suggested that self-reports be combined with technological approaches to gain a deeper understanding of the user experience. In future work, a questionnaire will be designed to assess user experience through emotions.