Unmasking the Twitterverse: Analyzing Sentiments towards DeFi

Abstract

This paper investigates netizens’ viewpoints on Decentralized Finance (DeFi) using emotion theory and lexicon sentiment analysis via machine learning. The data of 15,000 tweets on DeFi is gathered through automated web-scraping. Emotion score is evaluated through sentiment lexicon analysis and includes anger, anticipation, disgust, fear, joy, sadness, surprise, trust, and primary sentiments. The supervised machine learning reveals a score of 47,054 sentiments from 15,000 tweets, showing predominantly positive and trust sentiments in the sample. The positive sentiment may describe potential of Decentralized market. Meanwhile, trust emotion was indicative of the market’s response to the transparency and security of the DeFi system. This study contributes to theoretically explaining the implications of the DeFi phenomenon under the lens of emotion theory.