基于潜在语义索引的个性化搜索引擎的研究

Research of Personalized Search Engine Based on Latent Semantic Indexing

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上具有先进性,针对该模型中文档集用户兴趣有效性问题提出了建立用户个性词典的改进算法,提出了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系统方案,同时对系统中关键词的学习与提取、个性化计算的动态调整提出了若干有益方法。

Personalized search engine is a kind of new intelligent search engines that can learn the interests of users and use it to help users filtering information. Latent Semantic Indexing （ LSI ） is a new document retrieval model that has an advantage of word and text retrieved. LSI is improved by user＇ s characteristic dictionary in this paper. A personalized search engine based on improved LSI that can dynamically learn the interests of users is implemented. In the progress of using the improved LSI, several methods in distilling keywords and adjusting the personalization computing dynamically are proposed.