Font Size: a A A

Analysis Technique, Based On Data Mining Of Traditional Chinese Medicine Prescription

Posted on:2010-10-09Degree:MasterType:Thesis
Country:ChinaCandidate:H M MengFull Text:PDF
GTID:2208360275998642Subject:Pattern Recognition and Intelligent Systems
Abstract/Summary:PDF Full Text Request
Traditional Chinese Medicine (TCM) prescription contains abundant knowledge and experience, prescription mining is the important part of TCM data mining. In this thesis, multiple Data Mining methods were synthetically used to analyze the TCM prescription. First some priori knowledge for the prescription analysis is obtained by preprocessing and mining of TCM data. Second, we adopted a set of methods to get the collection of the effects of the prescription, namely, search basic formulas and medicine pairs contained in the prescription to get their effects at first, then analyze the effects of the single herbal medicine. After that, a clustering method was used on the effects collection to get the result of summarized effects of the prescription by Affinity Propagation(AP). In this paper, technologies and algorithms of above process are researched, and the automatic analysis system for prescription is built. Concrete work is as follows:(1) Preprocessing of original TCM data is researched, the TCM data is extracted and standardized based on standard vocabulary, and the dosage of herbal medicine is quantified by clustering and fuzzy algorithm.(2) Mine the effects association knowledge for prescription analysis by association rules algorithm, and the SimRank algorithm is applied to obtain the similarity between two herbal medicines.(3) Fundamental formulas matching and application modes matching of TCM for prescription is researched, which fully uses the priori knowledge to analyze the prescription in the view of the basic formulas and medicine pairs and so on.(4) Summarize the effects of the prescription. The effects of the single herbal medicine were filtered according to the relation of the effect and Guijing. By using AP clustering algorithm, the summarized the effects of prescription from a collection of effects resulted from the previous steps can be got, and the summarized results are made comparison under different conditions.(5) Build the TCM prescription automatic analysis system. The system is convenient to manage the priori knowledge of TCM and the results of the data mining, it's also able to set parameters for prescription analysis flexibly.
Keywords/Search Tags:Data Mining, Association Rule, SimRank, Affinity Propagation Clustering, Prescription Analysis
PDF Full Text Request
Related items