Since Web3 service does not rely on an authorization server, it uses the user's trustworthiness as a substitute for the authorization process. The user trustworthiness is based on transaction records within a community. On the other hand, it is difficult to derive the user's trustworthiness properly in the initial phase of community formation because there is little NFT (Non-Fungible Token) transaction data. The trust scoring system provides a general trust model based on GNN (Graph Neural Network) for an environment with a limited number of users. The trust model predicts user transactions and calculates trust scoring, which are then used for authorization in Web3 service. This trust scoring system is implemented in the authorization mechanism of Web3 service. The trust scoring system enables appropriate access control in all phases of community growth.
YASUNO et al. (Wed,) studied this question.