摘要
基于agent的信息检索系统在海量网络信息检索中已经得到广泛的应用。提出了一个基于多兴趣agent新的信息检索模型,该模型对查询信息进行了兴趣预测并生成了用户兴趣项权重向量。实验表明,模型在预测用户兴趣的效果和推荐搜索信息的准确率方面比传统的单兴趣Agent检索系统IRHO IA有一定的改善。
Agent-based information retrieval system has gained great attentions in huge web resource searching. In this paper, we will present an information retrieval model based on Multi Interest Agent. After making interest prediction on user＇ s query information, the model generated user＇ s interest item weighting vector. Experiment result shows that, compared with the traditional single interest Agent retrieval system IRHOMIA, it is more effective on deducing user＇ s interest and more accuracy in information recommendation.
出处
《计算机应用与软件》
CSCD
2009年第7期6-8,共3页
Computer Applications and Software
基金
国家自然科学基金项目(60702056)
