论文部分内容阅读
随着高校图书馆大数据时代的到来,读者有时很难找到自己喜欢的图书,会造成图书资源的浪费。针对这种情况,本文研究了基于用户分类的协同过滤算法在高校图书推荐中的应用,其中涉及读者分类、用户-项目评分矩阵的建立、向量空间模型的构建以及用户间相似度的计算,并考虑了高校图书和读者的特点,对用户-项目评分矩阵进行了改进,缓解了数据稀疏问题。研究结果表明,基于用户分类的协同过滤算法比传统的协同过滤算法计算复杂度低,在一定程度上优于传统的协同过滤算法。
With the arrival of the era of big data in college libraries, readers sometimes find it hard to find their favorite books, which will result in the waste of library resources. In view of this situation, this paper studies the application of collaborative filtering algorithm based on user classification in university book recommendation, including reader classification, user-item scoring matrix, vector space model construction and similarity calculation between users. Taking into account the characteristics of university books and readers, the user - project score matrix has been improved to alleviate the data sparseness problem. The research results show that the collaborative filtering algorithm based on user classification has lower computational complexity than the traditional collaborative filtering algorithm and is better than the traditional collaborative filtering algorithm to a certain extent.