基于三维交互效应的公共安全事件知识可视化研究

为提高公共安全协同管理水平与效率，提升措施与决策的针对性和高效性，本文提出对公共安全事件的分布特征及安全知识进行数据挖掘。选取时间、区域、性质维度进行频数统计，构建泊松对数线性模型将各维度的交互作用量化并通过知识可视化进行表达。研究表明：极大似然估计得到的交互效应参数作为公共安全事件维度交互的重要指标，知识可视化的直观展示能有效消除信息冗余并降低认知资源消耗，提高安全知识发现和信息资源认知的效率，从全新的角度为公共安全管理提供了科学可靠的指导依据。

Research on Knowledge Visualization of Public Safety EventsBased on Three-Dimensional Interaction Effect

In order to improve the collaborative management ability and efficiency of public safety, this paper proposes the data mining of the distribution characteristics and safety knowledge of public safety events to improve the pertinence and efficiency of measures and decision-making. The time, region and nature dimensions are selected for frequency statistics. A Poisson logarithmic linear model is constructed to quantify the interaction of each dimension, which is expressed by knowledge visualization. Research shows that the interaction parameter obtained by maximum likelihood estimation is an important indicator for the interaction of public safety event dimensions. The visual display of knowledge visualization can effectively eliminate information redundancy and reduce cognitive resource consumption, improve the efficiency of safety knowledge discovery and information resource cognition, and provide a scientific and reliable guidance basis for public safety management from a new perspective.