基于概念格的数字图书馆用户检索行为序列模式挖掘研究

Sequential Patterns Mining from Digital Library User's Retrieval Behavior Based on Concept Lattice

提出一种基于概念格的数字图书馆用户检索行为序列模式挖掘方法。该方法采用"基于概念格的自顶向下与分治相结合"的挖掘思想,通过自顶向下的概念格迭代,利用概念格的复用性和提取频繁项集的优势,获得数字图书馆用户检索行为的序列模式。该方法不需要遍历原始用户信息数据库,能够大大压缩挖掘时间,有助于数字图书馆提高用户检索速度、改进个性化服务。

This paper presents a method to mining the sequential patterns from the user＇ s retrieval behaviors of digital library based on concept lattice. This method searches out the sequential patterns of the user＇ s retrieval behaviors by mining ideas of ＂combining top - down and dividing - and - ruling based on concept lattice＂, using the application of the re - usability of concept lattice and its advantage in the extraction of frequent itemset. The method does not require comprehensive scanning to the original user information database, and it greatly reduces the time of mining that can help digital libraries to enhance the user retrieval speed, and improve the personalized services.