Research On Individualized Recommendation Method Of University Library Books Which Focuses On Science And Engineering

In recent years,the library collections in colleges and universities are constantly expanding.The number of books increased sharply,various kinds.Facing a large number of collection books,it takes a lot of time and energy for students to borrow books what they are interested in.How to borrow books more effectively becomes a problem that students care about.Therefore,personalized recommendation of books came into being.Personalized recommendation has been successfully applied to e-commerce,shopping and other fields.The user's behavioral preferences are obtained by processing,analyzing and discovering a large amount of user information.But the results of traditional recommendation methods are often not very satisfactory.The traditional collaborative filtering algorithm needs to be improved.After reading the relevant literature,combined with the data in this paper,an improved recommendation algorithm based on clustering and association rules is proposed.Recommend from the two perspectives of users and books to form the final recommendation set.On the one hand,from the user's perspective,users are clustered according to their borrowing habits.The similarity between users is then calculated within the class.User similarity calculation not only considers the similarity of traditional user borrowed book information,but also consider the user's own attributes to jointly build similarity calculations.The final user similarity matrix is obtained by summing the similarity of the two parts according to a certain proportion.On the other hand,starting from books,the relevance between book categories is constructed.The two parts are optimized to get the optimal recommendation results.Finally,the improved recommendation algorithm is compared with the traditional recommendation algorithm and the clustering recommendation algorithm,and the accuracy and recall rate of the three algorithms are compared.The experimental results show that the improved recommendation algorithm based on clustering and association rules has improved the accuracy and recall rate compared with the traditional recommendation algorithm,achieving the improved effect and purpose.