Applying a Social Reputation Network to Determine the Impact of Twitter Influencer Sentiment on NFT Market Activity

Abstract

Non-fungible tokens (NFTs) have exploded in popularity in the year 2021. The rise in mainstream attention towards this new asset class is largely attributed to influencers promoting their NFT investments to their followers across the most prominent social media platforms. These social media influencers, users who have established a reputation for their expertise on a given subject, make regular posts about that topic on their preferred platform, generating a significant following of avid supporters who engage with this published content. Social media users look up to these influencers to guide their decision making. For this thesis, we investigate how the transaction activity in the NFT market reflects the impact of influencers posting about NFTs on the Twitter platform. We introduce a framework for modeling NFT influencer’s reputations and leverage this reputation measurement system to determine the correlation between daily Twitter sentiment regarding specific NFT collections and their subsequent sale prices on the OpenSea NFT marketplace. We found that weighting our daily Twitter sentiment calculations with the reputation scores derived from our user-to-user reputation network significantly enhances the predictability of abnormal fluctuations in sale prices of the two NFT collections studied. Although the NFT market is immature, the results presented in this thesis suggest that further studies related to this field should consider the reliability of the individual content publishers when measuring the impact of social media sentiment on NFT transaction activity.