Food And Drug Safety Detection System Based On Massive Data Analysis

Today,the accidents of food and drugs' safety happen frequently.One reason is the huge development gap of inspection offices of various regions,independent storage of data and inconsistent data format standards which results in data isolation,difficult sharing of data resource.Another reason is that nobody taps the potential value of the test data in the post processing to advance warning.To solve these problems,this paper studies food and drugs' safety evalu,ation algorithm based on the fuzzy classification and standardizes data format of mass of food and drugs'inspection and obtains risk assessment level through calculation and analysis,so the issues of food and drugs' safely can be effectively monitored and early warnings.The main work of this paper is as follows:Firstly,the evaluation algorithm of food and drugs' safety based on fuzzy classification is studied.The quality level of the product is determined by pattern recognition with establishing the mathematical model of the detection data and calculating the membership degree of the comprehensive evaluation.By summing up and summarizing the proportion of continuous batch of inspections' number of various quality grade,setting up data visualization model for each grade of quality inspection,comparing trends of inferior product and combining with knowledge rules,comprehensive risk assessment results are obtained.Secondly,detection system of food and drugs' safety based on massive data analysis is developed which applies technologies such as cloud computing,Hadoop computing framework,HDFS distributed file system,the computing framework of Map/Reduce framework and HBase database to the detection system of food and drugs' safety.The system realizes the standardization of massive data format and solves the waste of space,load balancing,dynamic backing up,dynamic migration and other issues through the virtualization of storage server the and distributing of data resources,so that the system performance is stable,and has high reliability,scalability and good interactive interface.Thirdly,the experiment was carried out on the experimental data set of fermented milk.The data have 20 different brands of fermented milk and each brand has 20 batches of samples which have a total of 168 thousand samples.Through the analysis and calculation,we can get the risk level of these 20 brands.The experimental results show that the algorithm which has a certain degree of authenticity.The food and drugs' safety detection system designed in this paper which is based on the analysis of massive data,storing mass and various detection data effectively,realizes fast and accurate response of large-scale complex detection data query.The analysis of the existing data mining,assess the risk level of the goods detected provides data support for the food and drugs'safety early warning,so system has certain practical value.