面向政府决策需求的社会舆情信息语义组织研究

Semantic Analysis of Social Public Opinion Information for Government Decision Making

【目的】对海量异构的社会舆情信息进行语义组织,揭示社会舆情信息之间的多维度关联,以为后期的挖掘、分析和决策支持提供重要支撑。【方法】从政府决策需求出发,建立基于本体的社会舆情信息语义组织框架,根据社会舆情信息本体的特点,采用七步法的本体构建方法,以ABC事件本体为参考模型,明确其中的概念以及概念间的关系,建立社会舆情信息之间的语义关联。【结果】构建适用于社会舆情信息描述的POIOM模型,完成本体管理模块的设计。【局限】论文以访谈和统计分析为主要需求获取渠道,有待更多的政府决策者用户群体对需求加以验证和完善。【结论】POIOM模型对社会舆情信息具有良好的描述能力,从内容角度深层次地揭示同一案例不同特征项内容之间的关联,以及多舆情事件之间的逻辑关系和内在联系,基本验证了基于需求的社会舆情信息语义组织方法的可行性和有效性。

[Objective] This paper identifies multi-dimensional associations among social public opinion data with the help of semantic organization analysis technology. It also provides important support for data mining and decision-making tasks in the future. [Methods] We first studied the governmental decision-making needs to establish a semantic organization framework for the public opinion information. Second, we analyzed the characteristics of these information and then adopted the ＂seven-step-method＂ as well as the ABC event Ontology model to create a new one. Finally, we established the semantics connection among the public opinion information elements. [Results] We proposed a new POIOM model and management module for the public opinion information. [Limitations] This paper used interviews and statistical analysis to assess the information needs. More participants should be included in future studies. [Conclusions] The POIOM model could effectively describe the social public opinion information. It can also reveal the association among different features of the same case, as well as the logical and internal relationship among multiple public opinion events. This study shows that the semantic organization method is feasible and effective to process social public opinion information.