首先分析了数字图书馆中常用的用户知识挖掘技术——关联分析、分类/聚类分析、时序分析等,重点探讨了从数字图书馆访问日志中挖掘用户访问模式的方法和步骤,并以实验验证之。该方法有利于更好地理解用户的查询行为与需求,可用于获取和预测用户的动态知识。
This paper firstly analyses several user knowledge mining technology commonly used in digital library, such as association analysis, classification/clustering analysis and Sequence pattern analysis, and mainly discusses the methods and steps to mine user access patterns from the access logs in digital library. At last there is an experiment to test the methods. The finding of this study leads to a better understanding of user query behavior and demand, and can be used to acquire and forecast the user dynamic knowledge.