语义支撑下以原油市场为例的Web知识发现关键方法及实证研究

项目名称： 语义支撑下以原油市场为例的Web知识发现关键方法及实证研究

项目编号： No.71271200

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： 郭剑锋

作者单位： 中国科学院科技政策与管理科学研究所

项目金额： 54万元

中文摘要： Web成为在新视角下金融市场的知识库和传感器。近期，面向金融市场分析的Web知识发现成为急速升温的研究热点。但该交叉领域尚缺少全过程角度的系统的方法体系研究，并且在该领域中典型金融市场和大宗商品市场的个性尚没有被区分，且开源数据在大宗商品市场分析的应用尚没有提升到Web知识发现层面。为此，本项目以大宗商品市场为研究对象，用原油市场为例，基于语义技术，提出了面向市场分析知识的语义建模方法和基于知识单元的Web数据抽取与融合方法，来应对面向市场分析的Web知识发现过程中必然面临的知识圈定和Web数据有效获取问题。在此基础上，通过数学模型实验分析若干Web与原油市场间的基础性互动关系，验证方法研究部分的有效性并得到有益于实践的实验结果。本项目是对基于Web的金融市场分析方向的重大推动，在实证研究部分将发现和提出若干有普适性或有益于原油市场分析的基础性规律。因此，本项目具有学术和实践的双重意义。

中文关键词： Web；知识发现；石油市场；本体；联动

英文摘要： Under new view, Web has become the repositories and sensors for financial markets. Recently, the field about analyzing financial market based on Web data has become a research focus rapidly. But, the hot field has no systematic researches on method system from whole process view yet, and the differences between classical financial markets and commodity markets are not identified in this research field. In fact, the application of open source data in commodity markets does not update to the level of Web knowledge discovery yet. So we select commodity markets as our research objects, in the case of oil market. In this research, based on semantic techniques the method for semantically modeling for the knowledge of market analysis and the methods of Web data extraction and Web data fusion oriented to special knowledge unit are proposed, to deal with the issue of knowledge delineation and how to extract effective Web data we must encounter in the knowledge discovery process. Based on above methods we proposed, some empirical studies will be implemented to find the basic interactions between Web and oil market, which will check the validity of the methods we proposed, and contribute some useful empirical results for practice. The project wills strongly impulse the research field of financial markets analysis with Web,

英文关键词： Web；Knowledge Discovery；Oil Market；Ontology；Co-movement