为提高搜索引擎的个性化信息检索能力,通过构建个人兴趣搜索智能agent子系统SSPISIA来搜集、组织、挖掘和应用用户的个人兴趣信息。着重介绍了SSPISIA的实现,包括逻辑组成、学习方式、工作过程以及基于页面浏览时间和内容选择的个人兴趣度量规则,并在此基础上给出了基于SSPISIA数据收集的个人兴趣增量挖掘算法。实验表明该结构和算法不仅能够反映用户的长期兴趣,而且能够跟踪用户的短期兴趣变化,具有良好的适应性,进而为实现搜索引擎的个性化信息检索奠定了基础。
In order to improve the personalized information retrieval ability, the article constructed SSPISIA to collect, organize and mine users' personalized preference. Subsequently, discussed the logical organization, learning pattern, organization process and interest evaluation rules. At the same time, presented an algorithm of personal interest mining based on SSPISIA. Experiment shows that the architecture and algorithm aforementioned can trace user' s personal interest effectively and have excellent adaptability.