项目名称: 基于关联数据的信息聚合模型与实现研究
项目编号: No.71273225
项目类型: 面上项目
立项/批准年度: 2013
项目学科: 管理科学
项目作者: 潘有能
作者单位: 浙江大学
项目金额: 54万元
中文摘要: 关联数据可以被视为语义网的一种实现方式,它使得来自于不同来源的数据相互关联,从而促进了万维网的发展。目前已有的信息聚合技术主要面向传统的万维网,而基于关联数据的信息聚合技术主要面向语义网,关注的焦点在信息本身而不是承载信息的页面或是信息的访问接口API,聚合的对象直接深入到细粒度级的具有语义信息的实体。本项研究拟在对不同的数据集进行本体映射和数据链接的基础上,构建基于关联数据的信息聚合模型,并设计开发实验系统,在LOD数据集中进行测试和评估,实现对动态、分布和异构的数据源进行细粒度的语义信息聚合,为用户提供多角度、全方位、可视化的访问和了解信息的途径,在此过程中,语义相似度的计算及动态关联分析技术是待解决的关键问题。本项研究将遵循"理论-模型-实证"的技术路线,采用文献查阅、专家咨询、模型构建、系统设计、程序开发、实验结果分析等方法,从理论与实践相结合开展研究。
中文关键词: 关联数据;聚合;本体映射;数据链接;语义匹配
英文摘要: As an implement of Semantic Web, linked data connected data from different sources to accelerate the development of WWW. Be different from the existing information mashup technologies which mainly focus on the traditional WWW, the information mashup technologies based on linked data mainly face the Semantic Web. It focuses on the information itself instead of the pages with information or the APIs accessing information. The objects of mashup are semantic entities which belong to low-granularity level. Based on the ontology matching and data linking between different datasets, this research will build a model of information mashup which is based on linked data, and design an experiment system, then test and evaluate in LOD datasets. This research will realize low-granularity semantic information mashup to dynamic, distribute and heterogeneous data sources, providing an approach of multiangle, all dimensions and visualization to access and understand information. In this process, the calculation of semantic simmilarity and the dynamic linking analysis are the key problems to be resolve. This research will follow the technology roadmap of "theory - model - application", adopt methods which include document retrival, expert consultation, model building, system design, programming and result analysis. This research w
英文关键词: linked data;mashup;ontology matching;data linking;semantic alignment