The Research on User Interest Mining Based on Deep Learning

Abstract This paper, user behavior depth technology is studied, so that user information and search engine can interact intelligently. The system can get information that is conducive to mining users’potential search interests and preferences, so as to use this information to guide the search process and results filtering, and achieve efficient and accurate retrieval. The research process uses the mining of user logs of large data search engines, the establishment of user query classification system, and the deep mining of the behavior of user groups with different characteristics. User social network can be used to create a retrieval system in which users cooperate with each other because of their similar interests. This paper studies the calculation and method of user interest similarity to discover user groups with similar interest. User relationship network and system mining user relationship network are compared to promote the integration of the two and provide a basis for the application of recommendation system.