内容和结构化数据多维分析关键技术研究

项目名称： 内容和结构化数据多维分析关键技术研究

项目编号： No.61003051

项目类型： 青年科学基金项目

立项/批准年度： 2011

项目学科： 轻工业、手工业

项目作者： 彭朝晖

作者单位： 山东大学

项目金额： 7万元

中文摘要： 本项目研究内容和结构化数据多维分析的关键技术，旨在实现内容数据和结构化数据的深度集成分析，为用户全面洞察数据蕴藏的规律提供有力支持。主要研究内容和成果包括：（1）非结构化数据处理方面，提出一种基于SVM和扩展条件随机场的Web实体活动抽取方法，能够从Web上准确的抽取实体的活动信息，提出一种基于Markov逻辑网的两阶段数据冲突解决方法，考虑不同属性间冲突解决的相互影响，从而提高数据冲突解决的准确率。（2）结构化、非结构化数据关联方面，提出一种基于两层条件随机场的评论与数据库实体匹配方法，将相关评论信息与已集成的数据库实体相匹配，为用户提供更为全面的实体全景视图。（3）文本数据OLAP建模方法方面，提出了一种新的框架来将文本分析和OLAP进行结合，利用信息抽取和文本挖掘的方法来进行非结构化数据的多维分析，另外提出一种社会信息网络的多维数据分析方法，有效分析信息网络的信息。本项目已发表论文10篇，全部为EI收录，其中2篇论文分别发表在计算机学报和软件学报，项目成果预期将在商业智能、企业资源管理、情报分析等系统中得到广泛应用。

中文关键词： 内容数据；结构化数据；多维分析

英文摘要： The project examines the key technologies of multidimensional analysis of the content and structured data,so as to achieve the deep integration and analysis of the data, which strongly supports the user to insight into the hidden patterns in the data.The main research points include:(1)In the area of unstructured data processing, the project presents a method based on support vector machine (SVM) and extended condition random fields (ECRFs) to extract Web entity activities accurately, and a 2-stage approach for resolving data conflict based on Markov Logic Networks is proposed, in which mutual influence of different attributes are considered, so that the approach can resolve the integrated data conflict effectively and more accurately.(2)About the technology of semantic association between the content and structured data, a method based on 2-layer Conditional Random Fields(CRF) to match relevant reviews to integrated database objects is proposed, providing users the more complete holistic views of entities.(3)In the area of OLAP modeling on text data, a new framework combining text analysis and OLAP is proposed, which is based on information extraction and text mining and gives multidimensional anaylsis results well on unstructured data and social information network.Many important resuls have been obtained on theory and technology, 10 papers have been published and indexed by EI. It is expected that the achievements of the project are widely used in business intelligence, enterprise resource planning, and intelligence analysis.

英文关键词： content data; structured data; multidimensional analysis