融合文献多元特征的学术搜索策略优化研究  

Research on Optimization of Academic Search Strategy Based on Multiple Features of Literature

作　　者：梁柱[1] 汤斌 沈思 Liang Zhu

机构地区：[1]南京农业大学信息管理学院,江苏南京210095 [2]南京理工大学经济管理学院,江苏南京210094

出　　处：《情报理论与实践》2022年第7期29-35,17,共8页Information Studies:Theory & Application

基　　金：国家自然科学基金面上项目“基于深度学习的学术全文本知识图谱构建及检索研究”的成果,项目编号:71974094。

摘　　要：[目的/意义]在学术资源搜索领域中,用户存在使用语义宽泛的查询式进行搜索的行为,该行为导致搜索结果的冗余性,降低了用户的搜索效率。通过搜索结果多样化的方法,可以优化搜索结果排序,提高用户的搜索效率。[方法/过程]结合江苏省科学技术情报研究所的文献搜索平台日志数据,探讨了学术文献特有的分类组织方法在学术搜索中的作用,并在传统搜索结果多样化方法的基础上,提出了一种融合文献多元特征的学术搜索排序优化方法。[结果/结论]实验结果显示,学术文献特有的分类组织方法能在一定程度上提高用户搜索效果,而融合文献多元特征的排序优化方法,能有效提高用户的搜索效率。[Purpose/significance]In the field of academic search,the user has the behavior of using broad query to search,which leads to redundant search results and reduces the user’s search efficiency.This paper optimizes search results ranking and improves the search efficiency of users through search results diversification.[Method/process]Combined with the log data of the literature search platform of the Jiangsu Institute of Scientific and Technological Information,this paper discusses the role of the unique information organization method of academic literature.Based on the traditional method of Search Results diversification,an optimization method that integrates multiple features of documents is proposed for academic search ranking.[Result/conclusion]The experimental results show that the unique information organization method of academic literature can improve the search effect of users.On this basis,the optimization method of ranking that integrates multiple features of documents proposed in this paper can effectively improve the search efficiency of users.

关 键 词：学术搜索 搜索结果多样化 文献多元特征 策略优化 

分 类 号：G252.7[文化科学—图书馆学]