中文摘要：

本文提出一种基于图书内容的图书推荐算法。该算法利用词频-逆文本频率抽象图书特征向量,采用欧式距离度量图书相似性,使用CNM算法对图书相似性网络进行聚类,得到已知类别。当读者用户阅读、购买某本图书时,能够将该类别里的其他图书推荐给读者用户,方便其阅读或购买。

英文摘要：

A book recommendation algorithm based on content was proposed. The method used the term frequency-inverse document frequency to represent the books eigenvector,computed simi larities between books by the Euclidean distance, clustered booksvia the CNM algorithm , and finally obtained the classes of books. When users read or bought books, the metliod could recommend ot lier books in the same category. It is heljDfulfor users to read and buy books.