基于SNS的图书推荐系统研究与实践

摘要

In order to improve the quality of information services, personalised recommender system can be applied in digital library systems to help users find literature resources more conveniently and improve the utilisation of library resources. When facing the problem of data sparse, traditional recommendation algorithm has its limitations. To solve this, a book recommender system based on social network software is proposed and implemented. Through utilising three personalised book recommendation algorithms, the system is able to fully mine the user data, constructs user interest model more accurately and pushes better recommendation results to users. Data analyses results show that the introduction of additional social relational data can help to enhance the performance of the book recommendation system.%为提高信息服务质量,数字图书馆可利用推荐系统为用户寻找文献资源带来便利,并提高图书资源利用率.传统的推荐算法面临数据稀疏等问题时有其局限性.针对此问题,提出并实现一个基于社会网络软件的图书推荐系统,系统应用三种个性化图书推荐算法,能够充分挖掘用户数据,建立准确的用户兴趣模型并推送良好的图书推荐结果.数据分析结果表明,引入额外的社会关系数据有助于提升图书推荐系统的性能.