Research On The Core Prescription Discovery Algorithm Of Famous Veteran TCM Doctors For Lung Cancer Treatment Based On Complex Weight Networks And Association Rules

Traditional Chinese medicine(TCM)is the treasure of the Chinese traditional culture,which plays the important role in the society,especially for the treatment of malignant tumors.The famous veteran TCM doctors are leaders in TCM field.It is helpful to inherit and develop TCM facilities by analyzing their data of recipe.However,the research of using current information technology to analyze the data of the famous veteran TCM doctors'data has just begun.The main problem is that there are not sufficient data,which is hard to clean up.In this study,methods like focused crawlers,word segmentation tools,and regular expressions are mentioned to collect TCM data,which are stored in the TCM lung cancer prescription-drug database.The relationship between TCM prescriptions and medicines in prescriptions can be ed into complex network models.Based on this database,this study uses a complex network method to discover the core prescriptions in the prescriptions.Then,in this study,the association rule algorithm was used to study the drug relationship in core prescriptions and the dose-effect relationship between drugs.In summary,the innovation of this paper is reflected in the following three aspects:1)This paper proposes a multi-source heterogeneous data integration method that combines the focused crawler technology,word segmentation technology,regular expressions and data integration methods.A large dictionary of TCM drugs is constructed based on the commonly used doses of "Chinese Herbal Medicine" and clinical prescriptions,and it is used as the standard for inclusion in prescriptions to improve the quality of data integration.2)This paper proposes an algorithm for the discovery of core prescriptions of the famous veteran TCM doctors for treating lung cancer based on directed weighted complex networks.In the directed weighted network model,the initial weight of PageRank is set based on the drug's in-degree to obtain the core drugs.And then determines the core prescription based on the confidence of the core drug and prescription.3)In this paper,the Apriori algorithm was improved based on spectral clustering and information entropy to determine the association relationship between the drugs in the core prescription and the proportional relationship between the doses of the drugs.We can just input the information of the drug and convert it into a transaction matrix,which overcomes the defect that the traditional Apriori algorithm requires the researchers to complete the matrix conversion.Using spectral clustering to divide the prescription into several clusters,and using information entropy as the weight of the transaction,the characteristics of the transaction itself can be fully considered,and more association rules can be mined based on this.In addition,using drugs and their doses as items can also explore the dose-effect relationship between drugs.