关联规则算法在图书自动推荐系统中的应用

Research on the Book Automatic Recommendation System Based on Association Rules

针对传统图书自动推荐系统准确性不高的缺点,提出利用数据挖掘中的关联规则算法技术将读者借阅的图书、性别、年龄、职称、职业、受教育程度、爱好等多维关系生成关联规则,再将读者基本信息与这些规则进行比较,把匹配的关联规则推荐给读者,就能解决传统推荐系统的不足,提供更加灵活的个性化图书推荐服务。文章以湖南图书馆2011年读者借阅数据为例,利用Microsoft SQL Server 2008为工具进行了关联规则算法的数据挖掘分析。

Directed at the problem that traditional automatic book recommendation system has low accuracy, this thesis uses the method of association rule algorithm in data mining to generate the multi dimensional relations such as books borrowed by readers, readers＂ gender, age, title of the posts, occupation, education level and hobbies, into association rules. Then it compares the basic information of readers with those rules and recommends matched association rules to readers. In this way, it can make up the deficiencies of traditional recommendation system and provide more flexible and individualized book recommendation services. This thesis takes the reader borrowing data of Hunan Library in 2011 as an example and uses the tool of Microsoft SQL Server 2008 to conduct a data mining a nalysis of association rule algorithm.