基于情境推演的微博突发事件预测模型研究

MEPD:A Micro-blog Emergent Incident Prognosis Model Based on Situation Deduction

微博信息平台的高速发展,给网络突发事件管理带来了信息量激增等新挑战。为应对这些问题,研究并提出了一个新型的微博突发事件发现与预测模型,该模型基于情境推演算法构建,通过历史数据与现场情境数据的信息融合匹配进行事件发展预判;通过深层发掘和融合,对突发事件的未来情境进行预测和刻画。数据仿真实验证明,该模型具有较高的预测准确性、较全面的事件关键节点覆盖度和较好的时效性。

High-speed evolution of the micro-blog information platforms has brought many management challenges such as information deposition to Internet emergent incident managers. In response to the problems, a new micro-blog emergent incident prediction model was researched and proposed based on the situation deduction algorithms. The model fused fusion historical data and field event data to predict the emergent incident prospects. And it used the re- lated situation space to mine the microscopic situation and integrated the key event nodes to forecast the development trend of emergent incidents. Data simulation results show that the model has higher prediction accuracy, greater cov- erage and better timeliness than the old does so.