The evolution of studies on social media sentiment in the stock market: Insights from bibliometric analysis

Abstract

Social media sentiment applied in the stock market is extracted from social media platforms and researchers have grappled with the way it influences different stock market features like returns, trading volume and volatility. The growth in Twitter, StockTwits, WeChat and Sina-Weibo social media platforms has provided investors with convenient avenues for expressing their opinions about the stock market. We seek to examine the evolution of textual sentiment in the stock market over the past decade. We used co-citation, bibliographic coupling and co-occurrence analysis to provide an overview of the structure of social media sentiment within the stock market. The findings from the study show that the concept of social media sentiment as applied in the stock market is multidisciplinary. Most of the studies are found in the computer science and mathematical sciences domains with a few in the economics and finance domains. More recent studies are centred on ways and methods of extracting sentiment from social media as seen by the emergence of such author keywords like “Natural language processing”, “machine learning” and “deep learning” in the second half of the decade of the sample period used in the study. In summary, “social media sentiment” in the stock market has many avenues of expansion as seen by permeating different research domains like physics, mathematical sciences, computer science and finance. To the best of our knowledge, this is the first study to examine the evolution of social media sentiment using bibliometric analysis.