Design And Implementation Of Academic Library Recommendation System Based On Information Visualization And Data Mining

Today, with the rapid development of information technology, library information construction is pushing forward, at the same time the construction of digital libraries has largely enhanced the quality of service of the academic libraries. However, facing the growing collection of library books, readers are often unable to find the necessary information from the vast literature in a short time, therefore, the traditional literature search services have been difficult to meet the needs of users. How to provide readers with a personalized book recommendation service to help them quickly obtain the required information accurately has become an important research topic in digital library construction.Recommender system is a tool based on information filtering technology to provide users with personalized service, designed to solve the problem of information overload, and is a good complement of search engine. Recommender system applications for digital libraries can be used as a supplement for traditional library search service, proactively recommends books for readers. By analyzing the reader’s borrow records, recommender systems can understand user’s preference, and further provide personalized recommendations for users to meet the information needs of different individuals.This thesis was based on theories of information visualization and data mining, combined the knowledge of library science, and designed an academic library recommender system prototype. Based on historical borrow records in Exlibris Aleph library integrated management system, the author made engineering practices using recommendation algorithms such as collaborative filtering and content-based filtering. Based on Django framework, the system implemented functions such as borrow record synchronization, personalized recommendation, related books recommendation and book rankings using tools of D3.js library, amCharts chart component, jieba word segmentation tools, etc. This thesis described the requirement analysis, design, implementation and test of the recommender system in detail, and focused on the design of recommendation algorithm and the application of data visualization in data presentation. The author completed requirement analysis, high-level design, recommender algorithm design, detailed design, implementation and test of borrow record synchronization subsystem, recommender engine subsystem, frontend subsystem, and backend management system.The research results have some practical significance for the exploration of recommender system for academic libraries, the author hopes that this thesis can play a role in promoting the implementation and application of recommender systems in academic libraries.