Research On Association Rules Of Diabetes Medication Rules

Diabetes mellitus is a chronic disease with high prevalence.The number of people with diabetes of our country ranks the first in the world and diabetes prevalence rate higher than the world average.Traditional Chinese medicine(TCM)as the main body of traditional medicine,there are many doctors engaged in the study of traditional Chinese medicine in the treatment of diabetes since ancient times,but in the clinical diagnosis of TCM,it is required to regard people as a circulatory whole,from the overall situation to the local,increasing the difficulty of diagnosis,is not conducive to the exchange of academic ideas and clinical experience and research.In order to explore the law of using TCM to treat diabetes mellitus,the association rule technique is used to excavate the regulations between herbs and the rules between dialectical type and herbs,the research is of significance to enrich and develop the theory of pharmacology of traditional Chinese medical formulae and provide new ideas and references for clinical medicine.First of all,through the Python's Scrapy crawler framework,we obtained medical records of a number of doctors to treat diabetes from the national service platform for academic experience of famous TCM doctor,the medical records were then preprocessed and the data were processed respectively according to the different fields.For the prescription data of each medical case,Python's Jieba segmentation tools were used to get the name of singe herb,and then encoded the herb names which standardized processing for the non-standard herb names appeared,in order to enable the algorithm to process these data,these Secondly,for the bottleneck problem of classical association rules algorithm,an association rule algorithm based on Boolean matrix weighting is proposed,which reduces the times of scanning database and redundant candidate itemsets,and improves the memory usage efficiency and the execution efficiency of the algorithm,compared with the classical association rules algorithm,it can mine frequent itemsets effectively and efficiently.Finally,the association rule algorithm is used to run the processed data,and through frequency statistics and set different minimum support and minimum confidence,the rules are dig out.Through the data mining study of 540 medical cases of diabetes diagnosis and treatment of traditional Chinese medicine,the frequency of each herb,the relation rules among the herbs and the association rules between the dialectical types and herbs were mined from the data,and two,three and four core combinations were extracted.Through the efficacy analysis of dialectical types and the corresponding herbs,verify the medical data processing and analysis of the study on the validity of the mining of diabetic medication rule,it can provide certain medication guidance and reference for clinical diagnosis.