基于贝叶斯网络的传统餐饮业食品安全风险分析  

Food Safety Risk Analysis of Traditional Catering Business Based on Bayesian Network

作　　者：刘文燕 高齐圣[1] LIU Wenyan;GAO Qisheng(School of Economics,Qingdao University,Qingdao 266061,China)

机构地区：[1]青岛大学经济学院,山东青岛266061

出　　处：《复杂系统与复杂性科学》2024年第4期142-148,共7页Complex Systems and Complexity Science

基　　金：教育部人文社会科学研究规划基金(20YJA630018)。

摘　　要：为明确传统餐饮业食品安全问题的风险因素,从源头上找到治理问题的有效措施,首先通过构建故障树模型进行风险因素识别,利用故障树的最小割集和结构重要度对风险因素进行定性分析;其次将故障树模型转化成贝叶斯网络,利用实例数据进行结构学习和参数学习;利用BN的双向推理能力进行定量分析,最终从餐饮经营者的角度给出各类型问题的敏感因素。结果表明:人员管理发生问题的概率最大;培训不足是发生食品安全事件的重要原因。In order to clarify the risk factors of food safety issues in the traditional catering business and find effective measures to address issues from the source.The fault tree is constructed to identify risk factors,and qualitative analysis of risk factors is done by using the minimum cut set and structural importance.Moreover,the fault tree model is transformed into a Bayesian network,and final BN model is finished by structure learning and parameter learning based on case data.The two-way reasoning ability of BN is used to conduct quantitative analysis,and sensitive factors of various types of problems from the perspective of catering operators are finally obtained.The results indicate that personnel management is most likely to cause problems.Insufficient training is an important reason for food safety issues.

关 键 词：食品安全 贝叶斯网络 故障树 风险分析 

分 类 号：F280[经济管理—国民经济] TS201.6[轻工技术与工程—食品科学]