Throughout above similarity calculation, the thesis obtains maximum 20 similar users and input the information and similarity value of the 20 users into the database table UserSimilarity. The following task is to make book recommendations for every single user—namely to recommend books borrowed by neighbors to target users. After the definition of two-dimensional matrix, the thesis could obtain each reader’s book recommendation information by inputting recommendation information in book recommendation two-dimensional matrix RecommendBooks. For intensifying the recommendation results, the preferences of readers have been specifically sorted. By far, the user-based collaborative filtering results have been stored in the corresponding table. When users log in the system, the system will automatically discern user information and recommend relevant books.