Representation of Emotions through Social Media: A Systematic Review

Description

Page: 14-26

Shailjaa¹, Ishaan Sunsunwal², Sarath C. Jayadevan³, and Aparna Pandey⁴ (Department of Psychology, Faculty of Social Sciences, Jamia Millia Islamia, Jamia Nagar, New Delhi¹, Independent Researcher, New Delhi²,
Department of Psychology, School of Humanities, Social Science and Management, Indian Institute of Technology (IIT) Bhubaneswar, Jatni, Bhubaneswar, Odisha^3,4)

Emotional representation is the external manifestation of internal emotional states. With the advent of social media, people increasingly express emotions on virtual platforms. Although there is a plethora of research studying emotions in the context of social media, there is no comprehensive synthesis of this ever-growing body of knowledge. This review aims to close the research gap by consolidating existing knowledge that explores the ways and patterns through which individuals represent various emotional states online. A search across four databases (Web of Science, Scopus, PubMed, APA PsycNet) retrieved 357 articles, out of which 11 were included. The PRISMA 2020 Guidelines were followed in reporting the findings of the narrative synthesis. Studies have explored strategic self-presentation, online popularity, emotions expressed around certain topics and technological developments in emotion detection. Twitter was the most investigated platform. Sadness and anger were emotions examined the most, although there is a lack of a uniform theory of emotions in existing research. Sentiments followed by semantics and lexicon were the most frequently examined characteristics in textual emotional representations. Various tools, ranging from self-report to sophisticated AI models, were found to be employed in recognising emotions through textual posts and images, sidelining attention to video and audio content on social media. Sufficient research is dedicated to developing technology to detect psychological aspects of social media content. Findings are discussed in light of a growing virtual landscape of emotions posing various challenges to and suggestions for future interdisciplinary research in psychology and computer science.