Research On China’s Food Security Evaluation Index System And Evaluation Methods

Food security has a pivotal role in the overall security of the country and is regarded as the first priority in the governance of the country.China has made historic achievements in grain production and grain yields on its own.However,it is undeniable that China’s food security is still facing serious challenges: the sustainability of food production is constrained,the arable land area is decreasing,farmers’ motivation to grow food is decreasing,and seed security is facing a "stuck neck".Therefore,scientific and objective evaluation of China’s food security,summing up experience and preventing problems before they occur,is of great significance to ensure China’s food security.Based on the new situation of China’s food security and the new challenges it faces,this thesis analyzes China’s food security situation from two perspectives:horizontal panel data of provinces,autonomous regions and municipalities,and vertical national level time series data by constructing a food security evaluation index system to reflect the level of food security.Firstly,we crawled the news text data about China’s food security issue from the official website of the Ministry of Agriculture and Rural Affairs and the People’s Daily Online,built an LDA topic model to dig into the potential themes,analyzed the research hotspots of the food security issue,and provided a reference basis for constructing the evaluation index system in the later paper.Then,based on the results of the LDA topic model and the basic principles of constructing the index system,the food security evaluation index system of each province,autonomous region and municipality was established,the comprehensive scores of each sample in the panel data were obtained by global principal component analysis,and the comprehensive scores of each sample were summarized in chronological order,and the clustering of the panel data was realized by Ward method clustering and K-means clustering algorithm based on the comprehensive trend distance,and the K-means clustering algorithm with better effect was selected by using the silhouette coefficient to divide the food security level of 31 provinces,autonomous regions and municipalities into 4 clusters for horizontal comparison.In addition,the evaluation index system was established for country-level data,and the indicators were assigned by entropy method,improved CRITIC method,and optimal weighting method based on the maximization of deviation,and the distribution consistency and internal consistency criteria were used to judge the goodness of different weighting methods,and finally the optimal indicator weighting method was selected to calculate the comprehensive food security score to compare the food security level of different years at the national level longitudinally.When comparing the food security levels of 31 provinces,autonomous regions and municipalities horizontally,the results show that the food security levels of Heilongjiang,Henan,Shandong and Inner Mongolia are in the first tier,while the food security levels of Chongqing,Fujian,Ningxia,Beijing,Qinghai,Hainan,Shanghai,Tianjin and Tibet are in the fourth tier.For the country’s overall food security level,the overall food security score curve for each year shows that China’s food security status has improved in all years except for 2013 and 2016,when there was a small decline in food security level,and China’s food security level is generally on an upward trend.