A Study on Data Asset Trading and Sharing Based on Graph Convolutional Neural Networks

Data asset trading refers to the activities of using data as the trading object, and conducting data buying, selling, sharing, authorization, etc. through data markets or data platforms. The purpose of data asset trading is to realize the value, circulation and capitalization of data, and thus promote the innovative application and social benefit of data. Data asset trading involves multiple participants, such as data providers, data demanders, data intermediaries, data regulators, etc., who have complex relationships and influencing factors, such as trust, price, quality, privacy, security, etc. Neural network is a computational model that simulates the structure and function of human brain neurons. It can use a large amount of data to learn and infer, and thus solve and optimize complex problems. Neural network can effectively deal with nonlinear, high-dimensional, multimodal, dynamic changing data, and has strong generalization and adaptation abilities. Data asset trading based on neural network refers to using neural network to construct and optimize various links in data asset trading, such as data asset representation, evaluation, matching, recommendation, pricing, transaction, evaluation, etc. This paper aims at the problem of facing many indicators for data asset trading risk evaluation. It uses random forest algorithm [1] to reduce noise and optimize the preliminary indicators. It automatically analyzes the preliminary indicators and iterates with feature importance calculation. It selects the indicators that have greater impact on data asset trading risk as the data asset trading risk identification indicators. To address the problem of not fully utilizing the temporal features of data, it introduces temporal memory network to learn the long-term dependency relationship between long sequence data of time steps. It constructs and represents the complex and deeper functions between risk identification evaluation indicators and data asset trading riskiness.