Using a complex network-related algorithm

combined with the drug action and dose of "monarch

minister

assistant and guide" drugs in the prescription

a new algorithm based on the drug dose and effect of the core drug compatibility rule of the prescription was constructed. By constructing the prescription subnet

the core drugs and their weights were determined according to the drug action and dose in the prescriptions. The complex network model of the traditional Chinese medicine prescriptions was constructed based on the core weights of the drugs. The complex network label transfer algorithm was improved to mine the core drug compatibility rules of related prescriptions. Taking prescriptions and medications for treating postpartum affection by cold in Dictionary of Traditional Chinese Medicine Prescriptions（《中医方剂大辞典》） as an example

143 related prescriptions and 16 flavors of drugs with a frequency greater than 20

88 association rules with a support greater than 10

and a confidence greater than 80 were collected to construct a complex network model of postpartum stroke medications with 74 nodes and 1245 edges. Fifty-one complex clustering results were found using a complex network algorithm. Through analysis and research

it was found that the traditional Chinese medicine for postpartum affection by cold is mainly through removing wind and dampness

dispersing cold and analgesic

and nourishing qi and blood. At the same time

the improved label delivery algorithm is compared with traditional association rules and other methods. It is considered that improved complex network label delivery algorithm can obtain some new and more targeted drug combination results around the core drugs.