论文部分内容阅读
在综合国内学术信息检索服务的现状和现有理论方法研究的基础上,以检索词推荐为研究对象,构建基于文献特征项共现网络的学术信息检索词推荐模型。模型包括基础文献存储模块、文献特征项抽取模块、文献特征项共现网络预处理模块、基于特征项的文献检索模块及检索词服务前端5个部分。利用实验验证基于特征项的共现网络用于检索词推荐的可行性,结果表明推荐模型结果与各检索项的检索词更具有相关性,推荐质量较好。
Based on the status quo of domestic academic information retrieval service and the existing research on theoretical methods, this paper takes the recommendation of the search terms as the research object and constructs the recommendation model of the academic information retrieval term based on the co-occurrence network of document characteristic items. The model includes five parts: basic document storage module, document feature extraction module, document feature item co-occurrence network preprocessing module, document retrieval module based on feature item and search term service front end. The feasibility of using co-occurrence networks based on feature items to validate the word recommendation was verified by experiments. The results show that the recommended model results are more relevant to the search terms of each search term, and the recommendation quality is better.