基于协同过滤算法的个性化图书推荐系统研究

Research on Personalized Book Recommendation System Based on Collaborative Filtering Algorithm

针对高校数字图书馆对读者需求信息挖掘不足,无法主动为读者提供个性化图书推荐服务的问题。该文引入所罗门学习风格量表,多维度、全方位的构建读者特征模型,并提出基于随机算法与协同过滤推荐算法的混合推荐算法。首先,读者通过数据量表测试得到其学习风格,然后根据读者的浏览矩阵,在同种学习风格的用户群体中进行用户之间的相似度计算,最后采用Top-N的策略向用户进行图书推荐,为读者提供符合其个性特征的图书。实验结果表明,应用该算法能有效提高系统的推荐质量,达到良好的推荐效果。

Aiming at the problem of digital library in Colleges and universities lack of reader demand information mining, unable to provide personalized book recommendation service for readers. In this paper, we introduce the Solomon learning style scale,multiform dimension, construction of a full range of readers feature model, and puts forward a hybrid recommen- dation algorithm based on the random algorithm and user based collaborative filtering. Firstly, the reader by amount of data scale test get their learning style, then according to the readers browsing matrix, in the same learning style of the user groups of users between similarity calcu-lation, finally adapt the top-N strategies to recommend books to users, provide the reader with the book that satisfies their personalized need.Experimental results show that the proposed algo-rithm can effectively improve the quality of the recommend system and perform significantly better.