本项目(1)基于事件属性(文本关键词)关联语义链网络，通过文本标题核心语义，分而治之与逐步求精三种策略,对大规模Web文本时序信息集合进行事件识别与分类研究。在此基础上，(2)对特定事件与网站获取不同时刻的网民情感分布空间；(3)研究不同时刻事件网民情感分布空间的变化过程和事件属性、属性之间的关系、属性语义社区等三层不同事件语义粒度变化过程之间的内在关联机制，构建事件网民情感状态传播过程的基本机制与模型；(4)基于不同时刻情感分布空间的变化，通过“愤怒”与“难过”情感状态所占比重增加的速度与加速度形成的累积曲线曲率的变化，以判断网民情绪是否处于可控状态，继而构建网民情绪状态传播的基本机制与模型。实现上述四个算法软件web service的封装，供总集成平台与仿真平台调用。

The relationships among information, emotion and mood of web events are studied in the following steps. Firstly, the different web events are identified from the large-scale temporal text corpus on the web by three strategies: titles based core semantic discovery; divide and conquer; stepwise refinement. These web events form the information space. Secondly, the emotion tags on specified web events from web users are collected to construct the emotion space. Thirdly, the correlation between the temporal variation of emotion space and information space is discovered, which includes three different granularities: attributes, attribute relation patterns and semantic communities. Based on that, the emotion spreading process of web events is modeled. At last, the state of controllability of web users on a specified web event can be given by the temporal variations of emotion space, like the curvature variations of increasing speed and acceleration of ‘anger’ and ‘sadness’. Then, the mood spreading process of web events is modeled. Finally, the four parts will be wrapped as web services which can be called by the Integrated Platform.