Abstract: Some problems exist in the food safety network public opinion monitoring system, such as inaccurate data collection and low efficiency, unsystematic and inadequate data analysis, narrow data application and weak scene, so it is urgent to find out the solution. This paper summarized the current research status and development trend of food safety public opinion monitoring at home and abroad, sorted out the basic characteristics of food safety public opinion monitoring, and mastered the technical system involved in food safety public opinion monitoring, and proposed a solution to the problems of food safety public opinion monitoring in data collection, data analysis and data application. First, using keyword database to form monitoring probe and building dynamic media database would be beneficial to improve the efficiency and accuracy of data collection. Second, increasing the diversity of data analysis methods, such as public opinion classification, public opinion event processing and classification, multidimensional analysis, public opinion group analysis, public opinion disposal analysis, public opinion early warning analysis and policy support analysis, would be conducive to fully mining important public opinion information. Thirdly, the implementation of scenario-based services among the government, enterprises and consumers would be conducive to the innovative application of food safety public opinion monitoring.

Key words: online public opinion, monitoring, food safety, question, solution

摘要： 食品安全网络舆情监测系统存在数据采集不精准及效率低，数据分析不系统及不到位，数据应用面狭窄及场景化不强等难题，亟需探明解决方法。通过总结国内外食品安全舆情监测研究现况与发展趋势，梳理食品安全舆情监测基本特性，掌握食品安全舆情监测所涉及的技术体系，提出了我国食品安全舆情监测在数据采集、数据分析、数据应用三方面问题的解决方案。第一，借助关键词库形成监控探针、构建动态媒体库，有利于提高数据采集的效率和精准度；第二，增加数据分析方法的多样性，如舆情分级、舆情事件处理分类、多维度分析、舆情群体分析、舆情处置分析、舆情预警分析和政策支持分析等，有利于充分挖掘舆情重要信息；第三，落实政府、企业和消费者三方的场景化服务，有利于实现食品安全舆情监测的创新性应用。

关键词: 网络舆情, 监测, 食品安全, 问题, 解决方案