项目名称: 基于领域本体的产品质量安全Web信息分析
项目编号: No.71271013
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: 王理
作者单位: 北京航空航天大学
项目金额: 56万元
中文摘要: Web已成为产品质量安全信息发布、传播和舆情监测的重要平台,目前国内外基于Web的产品质量安全信息分析研究尚处于初期阶段,我国相关监管部门亟需这方面的理论和方法支持。通过对产品质量安全Web信息深入分析,可以及时、主动地对产品质量安全风险信息进行监控和预警。 本项目拟利用本体、信息抽取、Web挖掘等理论和方法,研究基于领域知识本体的事件框架模型,提出产品质量安全信息特征抽取算法;研究大规模产品质量安全领域文档共指消解、推理和实体链接方法;研究基于多源网络信息(如博客、微博、SNS)融合的评估方法,构建基于时序和地域的信息波动和扩散模型;最后,通过分析典型产品质量安全事件,建立产品质量安全Web信息生命周期模型,研究其传播路径、传播结构和演化机理。研究成果将揭示产品质量安全领域Web信息的产生、发展、演化和传播等内在规律,将为我国产品质量安全风险监控和预警提供理论基础、技术支撑和应用范例。
中文关键词: 领域本体;共指消解;影响力评估;传播规律;Web 信息
英文摘要: The Web has become an important platform of information bulletin, dissemination and monitoring public opinion for product quality safety. At home and abroad information analysis research for product quality and safety based on the Web is still in the initial stage, the relevant supervision department in China urgently needs the support of theory and method. The in-depth study the information for product quality safety, which is initiative and efficient for product quality safety monitor and early warn. This project intends to take advantage of ontology, information extraction, Web mining theory and technology, research on incident frame model based on domain ontology, build product quality safety ontology, propose attribute reduction algorithm and feature extraction method for product quality safety information; study the co-reference resolution, reasoning and physical link method for large-scale product quality and safety documents; research the comprehensive assessment model for the fusion multi-source(Blog, MiroBlog, SNS) web information and develop information fluctuations and diffusion model based on timing and geography. Finally, the project analyzes the typical product quality and safety incidents. The research results will reveal the inherent rules of information production, development, evolution and d
英文关键词: Domain Ontology;Coreference resolution;Evaluation of influence;Propagation;Web information