About the use of recommendation algorithm in this paper, recommendation algorithm is the collaborative filtering recommendation algorithm based on user and books. Similar users and similar books were explored. Similar users were defined as the ones who lent the same number of books, and the similar books were defined as the books which were borrowed together with the same times [10]. In the collaborative filtering based on user, the similarity between user v and user u was defined as n, the number of books borrowed together by v and u. The greater n was, the score vector between the user i and j was larger, and the greater the similarity cos(i, j) was, showing the two users were more similar. For each user u, the most similar k users were found, and books were recommend to according to the borrowing record of k users. In the collaborative filtering based on book, the similarity between book b and book t was defined as m, the number of they were borrowed together. The greater m was, the the greater the similarity they had. The first q books with the most similarities were found out to recommend to others. The user-based collaborative filtering mining algorithm is divided into three phases, namely, the establishment of user model, the search for nearest neighbor and the generation of recommended list.
About the use of recommendation algorithm in this paper, recommendation algorithm is the collaborative filtering recommendation algorithm based on user and books. Similar users and similar books were explored. Similar users were defined as the ones who lent the same number of books, and the similar books were defined as the books which were borrowed together with the same times [10]. In the collaborative filtering based on user, the similarity between user v and user u was defined as n, the number of books borrowed together by v and u. The greater n was, the score vector between the user i and j was larger, and the greater the similarity cos(i, j) was, showing the two users were more similar. For each user u, the most similar k users were found, and books were recommend to according to the borrowing record of k users. In the collaborative filtering based on book, the similarity between book b and book t was defined as m, the number of they were borrowed together. The greater m was, the the greater the similarity they had. The first q books with the most similarities were found out to recommend to others. The user-based collaborative filtering mining algorithm is divided into three phases, namely, the establishment of user model, the search for nearest neighbor and the generation of recommended list.
正在翻译中..