摘要
个性化搜索引擎是一种通过机器主动学习用户兴趣,并根据用户兴趣帮助用户进行信息筛选的新一代智能化搜索引擎,潜在语义索引模型在词与词、文本与文本之间的检索中具有先进性.针对该模型中文档集用户兴趣有效性低的问题引入用户个性词典来改进,给出一个完整的可学习用户兴趣并可动态调整的个性化搜索引擎的设计.实验表明潜在语义索引比向量空间模型具有更好的信息检索性能,同时改进的潜在语义索引算法与传统算法相比在文献检索方面性能有明显提高.
Personalized search engine is a kind of new intelligent search engine that can learn user＇s interests and help user to filter information. Latent Semantic Indexing（LSI） is a new retrieval model based on document that has an advantage of word and text retrievel. LSI is improved by user＇s personalized dictionary in this paper. A personalized search engine based on the improved LSI that can dynamically learn user＇s interests is implemented. The improved LSI is proved to perform better than VSM in the experiment and the improved LSI has a better performance than the traditional algorithm in the information retrieval.
出处
《苏州市职业大学学报》
2010年第2期54-57,共4页
Journal of Suzhou Vocational University
基金
健雄职业技术学院教改资助项目(200905)
健雄职业技术学院精品课程资助项目(200803)
关键词
个性化
潜在语义索引
向量空间模型
用户个性词典
信息检索
personalized
latent semantic indexing
vector space models user＇s personalized dictionary
information retrieval
