Research on User Access Pattern Mining Based on Web Log
Ting‐Ting Wu · Asia-pacific Journal of Convergent Research Interchange · 2020
Aiming at the field of e-commerce, this paper analyzes e-commerce user behavior based on the data characteristics of e-commerce back-end logs and constructs a user behavior mining model.On the basis of Web user behavior theory, it analyzes user behavior based on interactive content.Based on the background of big data, the traditional data mining algorithm is further optimized, which greatly improves the operating efficiency of the algorithm.At the same time, a distributed file storage structure is adopted to improve the fault tolerance of system data processing.This paper studies the advantages and disadvantages of collaborative filtering recommendation algorithms.The Web user behavior mining system constructed in this paper can conduct multi-dimensional and efficient mining.It helps e-commerce merchants and content providers to understand their users and achieve better commercial value through precise marketing and accurate recommendations, and complete the upgrade of data-driven services.