针对近年来频繁发生的食品安全事件,提出从时间要素和地区分布情况两个方面对食品安全事件进行统计分析,探寻其中隐含的内在规律,为有关部门对食品安全事件能够做出有效的预防和及时的应急响应工作提供充分可靠的数据支持。首先,将所涉及的食品安全数据进行预处理,构建面向食品安全突发事件的时间命名实体,选择条件随机场模型对时间表达式进行识别。从各地区食品安全事件随时间变化的角度,绘制并分析了事件高发地区的事件发生趋势图,根据趋势图预测未来事件发生的概率。然后,在地区分布情况方面,分别从单地区和多地区两个角度,统计并分析了各地区食品安全突发事件的概率分布情况。单地区包括频度、密度、广度三个指标,多地区包括共现频率、共现广度、点互信息和条件概率。结果表明:对食品安全事件时间要素和地区分布两个方面的统计和分析,可以有效地掌握食品安全事件的发生规律,并预测未来走向,为食品安全突发事件的预防、预测、预警工作提供可靠的数据支持。
<<Aiming at the frequent occurrence of food safety incidents in recent years,statistics and analysis of food safety incidents from two aspects of time factors and regional distribution are put forward,explore the hidden internal laws,so that people have a better understanding of food safety incidents,provide sufficient and reliable data support for the departments concerned to make effective prevention and timely response to food safety incidents. First,the food safety data were preprocessed to build a time naming entity for food safety emergencies. The time expression is identified with the conditional random field model. From the perspective of time change of food safety events in various regions,the trend map of event occurrence in high incidence area is drawn and analyzed,and the probability of future events is predicted according to the trend map. Then,in the area distribution,the probability distribution of food safety emergencies in each region is counted and analyzed from two angles of the single region and the multi region. The single region includes three indexes:frequency,density and breadth and the multi region include co occurrence frequency,co occurrence breadth,point mutual information and conditional probability. The results show that the statistics and analysis of two aspects of the time factors and regional distribution of food safety events can effectively grasp the law of food safety events and predict the future trend,and provide reliable support for the prevention,prediction and early warning of food safety emergencies.
<<