食品安全数据分析可视化模型研究

Research on visualization model of food safety data analysis

中文关键词:  食品安全数据  大数据分析  可视化模型  靶向监管

英文关键词:food safety data  big data analysis  visualization model  targeted supervision

中文摘要:

      目的 全面提升食品安全风险的感知、筛查、分级、预警和防控能力, 完善食品安全监管与治理体系, 客观正确的引导舆情。方法 采用大数据分析技术研究食品安全检测数据, 建立食品安全分析可视化模型。运用大数据分析平台Hadoop, 结合Tree Ensemble和Model-Based Ranking算法特征性分析食品抽检数据, 搭建BP神经网络, 结合Apriori和FP-growth关联分析等技术, 深度挖掘相关信息, 有效集成为对食品安全治理体系有价值的信息资源。结果 以折线图、热力地图等可视化模型实现食品类别、检测项目、结果分析、生产地址、检测数据分析及发展趋势等相关信息进行在线展示与分析, 能够满足关注食品安全的各领域人员很直观地就获得自己需要的食品安全信息的需求。结论 可视化模型能够增强数据分析结果的视觉效果和直接性, 切实提高食品安全风险预警的靶向性、有效性。

英文摘要:

      Objective To comprehensively improve the ability of food safety risk perception, screening, grading, early warning and prevention and control, improve the food safety supervision and governance system, and guide public opinion objectively and correctly. Methods Big data analysis technology is used to study food safety inspection data and establish a visual model for food safety analysis. Using the big data analysis platform Hadoop, combined with Tree Ensemble and Model-Based Ranking algorithm to characterize food sampling data. BP neural network is built, combining Apriori and FP-growth correlation analysis and other technologies to deeply dig relevant information and effectively integrate it into valuable information resources for food safety management system. Results The online display and analysis of food categories, testing items, result analysis, production address, testing data analysis and development trend could be realized by visual models such as line chart and thermal map, which could meet the needs of people in all fields concerned about food safety to obtain their own food safety information intuitively. Conclusion Visual model can enhance the visual effect and directness of data analysis results, and improve the targeting and effectiveness of food safety risk warning.