Impatto dei social media sui mercati NFT

The advent of Non-Fungible Tokens (NFTs) has led to the emergence of a socio-economic phenomenon of significant magnitude in terms of media attention and the prices at which these digital assets can be traded. Since these digital art assets lack inherent value, their sale prices are primarily influenced by the public attention they receive on social media and the surrounding hype. The objective of this research thesis is to characterize the socio-demographic attributes of users discussing NFTs on social media. Then, we aim to establish a connection between the discussion of NFTs on social media and the markets on which these objects are traded, exploring correlations between their popularity and sale prices. The research methodology consists of an initial data collection phase from a popular social media platform, Reddit, to obtain information on NFT discussions, and from blockchain markets (Ethereum, WAX) to gather information on which NFTs are being traded, including price and other metadata. Then, we attempt to link the two datasets through a matching process between the images posted on the social media platform and the corresponding images of NFTs traded on the market. Given the computational complexity of this process, the final phase of the project utilizes a pre-trained deep convolutional neural network called AlexNet for the conversion of images in a lightweight vectorial representation and employs a locality-sensitive hashing (LSH) technique to significantly speed up the matching of images using their feature vectors. Having established a fundamental connection between the social media dataset and the NFT markets dataset, we are able to open new avenues for this research project thanks to the potential offered by the combination of AlexNet and LSH, applicable only in the case of NFT markets due to the visual nature of these digital assets. These applications include the exploration of correlations between the popularity of an NFT and prices at which it is traded on Marketplaces, rather than finding coordination phenomena between users on social media who try to artificially manipulate prices on Marketplaces.