基于中图分类法的用户兴趣模型研究  

A RESEARCH ON USER PROFILE MODEL BASED ON CHINESE LIBRARY CLASSIFICATION

作　　者：邵志峰[1] 李荣陆[1] 胡运发[1] 

机构地区：[1]复旦大学计算机与信息技术系,上海200433

出　　处：《计算机应用与软件》2007年第8期85-86,108,共3页Computer Applications and Software

摘　　要：提出一种基于中图分类法的用户兴趣模型,形式化地描述了用户兴趣模型的建立和学习过程。在建立用户兴趣模型时,需要对代表用户兴趣的中图分类号进行挖掘,由于传统的Apriori数据挖掘算法更适合于处理无序的集合,而中图分类号中的号码是有序的。提出了一种改进的算法来自动构建用户兴趣模型,并据此开发了一个科技文献过滤系统作为典型应用。A user profile model based on the Chinese library classification is proposed and the construction and update process of this model are developed. As the traditional Apriori algorithm handles the unordered item sets efficiently, while the item sets in the chinese library classification are ordered, an improved Apriori algorithm is proposed to create the initial user profile model composed of the item sets automatically,and then the model and the algorithm are integrated into a scientific literature filter system as a representative application.

关 键 词：信息过滤 数据挖掘 用户兴趣模型 中图分类法 

分 类 号：TP393.4[自动化与计算机技术—计算机应用技术]