Finally, the book recommendation li

Finally, the book recommendation list was generated. The recommended results were kept in the database table, involving the reader’s name, library card number and the name of the recommended book. When readers log in the book recommendation system, the latter presented to the targeted recommended information to the readers according to the information of the readers and realized the personalized book recommendation. The principle of book-based collaborative filtering recommendation algorithm is similar to that of user-based collaborative filtering recommendation algorithm, and the major process includes the calculation of book similarity and the generation of recommendation. For the first stage, the similarity between books and books was computed. The basic idea of calculating similarity between book i and book j was based on reader-book evaluation matrix; the scores of each project was regarded as a n dimensional user space vector. Then, the similarity between the two was calculated according to the similarity measure formula. For the second stage, the first m (such as 20) books with the most similarity were selected according to the generated similar book list, and they were stored in the database table so that the book recommendation system could call at any time. The key step of book-based collaborative filtering algorithm was also to compute the similarity between projects and select the most similar projects, being similar to user-based collaborative filtering.