项目名称: 语义Web知识库补全关键技术研究
项目编号: No.61772079
项目类型: 面上项目
立项/批准年度: 2018
项目学科: 其他
项目作者: 王志春
作者单位: 北京师范大学
项目金额: 15万元
中文摘要: 大规模语义Web知识库是实现数据共享和构建智能应用的重要基础,知识库补全技术可以基于已有事实推断获得缺失事实,对于扩大知识库规模和提高其数据质量具有重要意义。已有的知识库补全方法只关注知识库中的关系型事实,而忽略了语义Web知识库普遍包含的本体以及非关系型事实。针对语义Web知识库的特点,本课题拟有针对性地研究:(1)基于本体约束的知识库补全方法;(2)关系型与非关系型事实的联合推断模型。本课题的研究将突破现有方法的局限与不足,在语义Web知识库补全问题上获得更好的效果。
中文关键词: 知识库;语义Web;知识库补全;本体;表示学习
英文摘要: Large-scale knowledge bases in Semantic Web form the important basis for sharing data and building intelligent applications; knowledge base completion technique can infer missing facts in knowledge bases based on the existing facts, which is very important for expanding and improving knowledge bases. The existing knowledge base completion methods only focus on relational facts, ignoring the ontology and non-relational facts in knowledge bases of Semantic Web. To take full advantage of these new features of knowledge bases in Semantic Web, this project will study: (1) ontology-based knowledge base completion method; (2) united inference model for both relational and non-relational facts. The methods studied in this project will overcome the limits of the existing methods, and will perform better in the problem of knowledge base completion in Semantic Web.
英文关键词: Knowledge Base;Semantic Web;Knowledge Base Completion;Ontology;Representation Learning