Abstract :
As online education continues to expand, understanding learners’ emotional experiences has become essential to enhancing digital learning environments. This systematic review investigates how emotion recognition has been studied in online learning contexts between 2010 and 2024, drawing on 17 peer-reviewed articles sourced from IEEE Xplore and ScienceDirect. The review addresses six key research questions. After a Boolean search string was used to gather most relevant articles from databases, several filters were applied. The selection of studies was based on strict inclusion and exclusion criteria: only studies written in English and published in Q1 or Q2 journals between 2010 and 2024 were considered, while systematic reviews or studies lacking relevance to emotion recognition in online learning were excluded. As a result, only 17 studies met the final criteria, highlighting the scarcity of research in this area. Qualitative coding of the selected studies revealed that discrete emotions such as happiness, surprise, anger, fear, and sadness were the most frequently examined, with happiness being the most commonly investigated. The findings also highlight the positive potential of emotion recognition technologies to enhance learner engagement and improve e-learning platforms, though challenges remain regarding participant diversity, data quality, and differences in educational settings. Notably, the limited number of studies, especially concerning language learning and varied populations, underscores significant knowledge gaps. This review offers valuable insights for researchers, educators, and policymakers aiming to develop more adaptive and inclusive online learning systems while emphasizing the need for broader, more diverse research across multiple databases and educational contexts.
