A study of machine learning techniques on non-fungible tokens

• Main Author: Tan, Ryan Yu Xiang
• Other Authors: Vun Chan Hua, Nicholas
• Format: Final Year Project
• Language: English
• Published: Nanyang Technological University 2023
• Subjects: Engineering::Computer science and engineering
• Online Access: https://hdl.handle.net/10356/166984
• Institution: Nanyang Technological University

The rise in popularity of Blockchain technologies and digital currencies has sparked an increased interest in Non-fungible Tokens (NFT). The ownership of an NFT can be proven by showcasing an immutable public transaction on the blockchain. As the variety and diversity of NFTs increases exponentially, there will be an information overload for new NFT owners who wish to make their first purchase. Creating a recommender system can shortlist the various collections that are well-received by other users based on the owner’s preferences or NFT attribute details. In this report, we will review existing literature on Machine Learning Techniques for creating recommender systems, as well as identify current research findings surrounding NFT. This report then presents the project which explores using Content-based Filtering and Collaborative Filtering techniques to suggest NFT collections that are worthy to invest. The project seeks to better understand if existing methodologies and techniques are suitable to be used in the NFT domain, since there are fundamental differences between NFT collections and items as such movies or music. The report also presents some of the challenges faced during this research project and provides some recommendations to improve future research in this domain.