基于VSM和偏好本体的个性化信息检索技术的研究

Research of Personalized Information Retrieval Technology Based on VSM and Profile Ontology

信息检索技术特别注重向量空间模型和偏好本体的结合。为便于找出用户输入的关键词间的关系，利用本体的关联分析及模糊本体值计算关键词间的相似度，本文采用加权算法和排序算法计算每个文档的权重，并根据文档权重进行检索结果的重排。为优化重排模型，本文还对每个检索对象的驻留时间进行合并，并使用衡量搜索引擎质量指标和评价标准F-measure对本文提出的重排机制性能进行测试。实验结果表明，使用本方法进行个性化信息检索的性能优于Google方法。

Information Retrieval（IR） techniques specifically focus on combination of Vector Space Model （VSM） with Profile Ontology. In this. paper, we propose a novel hybridization of the IR processing to calculate the weight of each document and to find relatiors between the user entered terms by using the weighting algorithm and the ranking algorithm, and takes advantage of ontology-based con＇elation analysis which uses the fuzzy ontology value to calculate the similarity score between terms and includes the re-ranking algorithms to display the search results according to the weight of the document. We incorporate the Dwell Time of each retrieval session to optimize re-ranked model, and the performance of our re-ranking mechanism using ]Discounted Cumulative Gain （DCG） and F-measure was tested. The experimental result shows that the Web retrieval efficiency achieves improvement when our personalized retrieval approach is compared with the Google search.