摘要
基于Lucene向量空间模型搜索的排序算法缺乏对自然语言语义理解的能力,直接有效的方法是根据用户个体对搜索文档的喜好,对选中的文档得分加权,由此提出Download-through Rank算法,对原有的排序算法进行了改进,设计并实现了个性化搜索引擎。实验证明,改进后的搜索排序算法能够有效提高信息检索的准确度。
The ranking pages algorithm based on Lucene VSM lacks the ability to understand natural language. The direct and effective method is to weight the selected document accormng to me user s individual preferences. It proposed an approach to improve the original ranking algorithm by Download-through Rank algorithm, and designed personalized search engine. Experimental results showed that the proposed sorting algorithm can improve the accuracy of the information retrieval.
出处
《蚌埠学院学报》
2015年第5期34-38,共5页
Journal of Bengbu University
基金
安徽工程大学青年基金项目(2013YQ29)
