摘要
[目的 /意义]构建基于关联数据的探索式检索系统,充分利用关联数据中隐藏的知识网络,向用户提供发现新知识的机会。[方法 /过程]基于DBpedia电影数据集,采用改进的向量空间模型进行关联数据的语义相似度计算,使用可视化的交互技术对检索结果进行呈现。[结果 /结论]通过基于任务的评价方法与IMDB进行对比,证明基于关联数据的探索式检索系统能够提高检索效率,提升用户体验并能引导用户发现其感兴趣的信息。
[ Purpose/significance ] Building a LOD-baesd exploratory search system can make full use of the hidden knowledge behind the linked data, and then can provide users with new opportunities to discover new knowledge. [ Method/process] Based on DBpedia movie dataset, we use improved Vector Space Model to calculate the semantic similarity and use visual interaction to show the search results. [ Result/conclusion] By task-based evalution, comparing with IM- DB, we prove that LOD-based exploratory search system can improve the efficiency of search and improve the user experi- ence of search and also can guide the users to find the information they are interested in.
出处
《图书情报工作》
CSSCI
北大核心
2017年第5期117-124,共8页
Library and Information Service
关键词
关联数据
探索式检索
语义相似度
linked data
exploratory search
semantic similarity