Study On The Food Safety Evaluation Based On Comprehensive Index Method And Rough Set Theory

Food safety is the root of people’s life,the foundation of national stability and the premise of social development.After decades of development,Chinese food safety laws and regulations,supervision and management,standard construction and risk assessment have become increasingly mature,but the evaluation of food safety science progress slowly.Food safety big data has been widely applied with the development of big data theory and statistical methods.More informations which are dugging deeply through statistical methods on the data resources of food detection are significance for scientific evaluation of food safety.Therefore,the paper applied comprehensive index method and rough set theory statistical method to study on food safety in China based on a large number of food safety testing data.The first part was the introduction,covering the significance of the topic,research content,research methods and innovative points.The second part was overview on food safety in domestic and overseas,the status and progress of food safety laws and regulations,supervision,risk assessment,standard system and standard evaluation were discussed.The third part were concepts and basic theories,indicating the concepts and theories related to food,food safety and food safety standards,elaborating the effect of food safety on the nation and society.The fourth part was food safety standard indicators system and evaluation method,introducing the current situation of food safety standard system in China,expounding and comparing the food safety evaluation methods,and proposing the food safety composite index algorithm and rough set theory algorithm.The fifth part was food safety evaluation based on comprehensive index method,comprehensively evaluating the food quality and safety of the 10 major categories(26 minor categories)and the effects of various inspection indicators and each single inspection indicators.The sixth part was food safety evaluation based on rough reduction,the rough set theory which was assist in Rosetta software was applied to evaluate the reduction,risk indicators and risk level of 10 major categories(26 minor categories)food standards.The seventh part covered conclusions and reconmmendations,involving the summary of the paper and proposals on improving food safety.The main conclusions were as follows:Firstly,the comprehensive index method is more suitable for analysis on big data of food quality and safety than the traditional qualified rate evaluation method.The total quality and safety of the main food were good,the safety degree were very high,the levels of 10 categories food were all better or above levels,soybean products were best level;pastries,grain processed products,beverages,edible oil,cooked meat products and spices were better level;wine,infant formula and liquid milk were good level.From 2012 to 2015,the overall quality and safety performance of Chinese major food products was stable,with 50% of the increase and decrease,respectively.The products which quality and safety were increased were pastry,beverages,grain processed products,edible oil and liquid milk;the reduced products were soybean products,cooked meat products,spices,wine and infant formula.In terms of subdivided food,25 of the 26 subcategories food had reached the level of better or higher,but the quality and safety evaluation of sterilized milk was general,and 17 subcategories of food have shown a downward trend.Secondly,Exogenous risks such as microorganism and food additive were not the main factors affecting food quality and safety in China.The two kinds of indexes had relatively little impact compared with the conventional physicochemical indicators.The conventional physicochemical indicators which had a great influence relatively on liquid milk,cooked meat products,wine and spices.The microbial indicators had a great influence relatively on pastries,beverages and soybean products.The indicators of food additives had a great influence relatively on liquid milk and spices.Each individual indicators had great difference on food quality and safety,there were 1 indicators with slightly polluting level,95 indicators with qualified level,50 indicators with middle level,36 indicators with better level and 325 indicators with best level.The indicators of slightly polluting level,qualified level and middle level had relatively large impact on food quality and safety,while the indicators of better level and best level relatively small impact,therefore the monitoring frequency should be adjusted appropriately.Thirdly,the Rosetta software of rough set had a good coverage of food monitoring indicators evaluation with an average coverage of 91.2%,a high prediction accuracy with an average prediction accuracy of 97.6%.The Chinese food standard system needs further optimization,the number of core indicators were relatively small.The 507 inspection indicators of the 10 major food standards were reduced to 89 core indicators and 418 general indicators.There were 41 risk indicators which suggested different degrees of risk were obtained from the evaluation,the risk monitoring should be strengthened.The risk indicators were as follows: The fermented soybean products were benzoic acid,the non-fermented soybean products were sorbic acid,benzoic acid,dehydroacetic acid,total number of colonies and coliform bacteria.The pastries were potassium sulphate and total number of colonies.The tea drinks were total number of colonies and tea polyphenol.The fruit and vegetable juice was total number of colonies.The drinking water were the total number of colonies,nitrite,total oxygen consumption and bromate.The fermented milk was fat.The edible oil were acid and benzopyrene.The soy sauce were benzoic acid,amino acid nitrogen and total number of colonies.The sauce were salt,amino acid nitrogen and benzoic acid.The vinegar was total number of colonies.The other spices were the total number of colonies,total ash,acid insoluble ash and peroxide value.The liquor were total acid,total ester,solids,ethyl acetate and ethyl caproate.The infant formula were vitamin A.The old infant and young children formula were protein,iron,selenium and calcium.The cooked meat products were dehydroacetic acid and total number of colonies.The innovation of this paper lies in: First,the paper improved the synthetic index method,the comprehensive index evaluation method which was applicable to the data of food inspection indicators,based on studying the limited method of food standard.Second,the improved comprehensive index method which based on the existing large number of food safety testing data was applied to comprehensively evaluate the quality and safety of major food in China.Third,rough set theory was applied for evaluating the indexes of Chinese main food standard system,including reduction,risk indicators and risk level.Fourth,the comprehensive index method and rough set theory reduction method was combined with using the consistency of evaluation index grading standard to evaluate food safety big data.