Research And Implementation Of An Expert System On TCM Clinic Acupuncture Treatments

TCM acupuncture as an important composed part of traditional Chinese medicine is theaccumulation of medical experiences and theoretical knowledge of Chinese nation on long time,and it is the treasure of the traditional medicine of Chinese nation. It is one of the fields mostworth to be researched, explored and practiced to reveal the inherent laws among the clinicaltreatments and provide basis of correlation, rationality and scientific theory of TCM acupunctureclinical treatmens that how to apply the theories, techniques and methods of expert systemresearches, designs and realizes an expert system of TCM acupuncture, and the relatedresearch reports are difficult to be found yet.A technical solution of TCM acupuncture expert system had been mainly explored,researched and designed in this paper, the system had been implemented by extending of thefunctions of the TCM acupuncture information database system based on the Java EE softwareframework, B/S architecture pattern and3NF and database standard.The primary work researched and developed and the original innovation points aresummarized as the followings:1. Implementation of a TCM acupuncture expert system. The system mainly consists of theknowledge library, the inference engine, the interpreter, the user interface and the extendeddatabase, in which the knowledge facts in the knowledge library are obtained after analysis bydata clearing, data mapping preparing and a K-means cluster algorithm running on the actual dataabout summarized information of a large number of acupuncture treatment literatures from theTCM acupuncture information database system.2. An inference process by applying a tolerance value. An inference mechanism based on atolerance value had been proposed in this paper, treatment solutions of exactness or approximaterange matched by the limitation of a tolerance value can be fed back to user. The inference processallows user to input a various of inquiring condition and feeds back a rich matched inquired resultcombined a tolerance value through comprehensive considering the characteristics of theknowledge facts after clustered. 3. An Improving clustering analysis algorithm. Mapping acupoints and treatment approachesof samples on to binary strings, an improved distance function and a kind of K-means clusteralgorithm with convergent condition through a tolerance value had been proposed in this paper,and four technical problems faced by traditional cluster analysis had been more well solved, i.e.,choosing a K value of classification number in advance, determining initial central sample,studying-self of central sample in process of clustering and a convergence standard of cluster.4.A strategy and its implementation approach to evaluate a balance point of design for clusteralgorithm effect. A feasible approach had been provided in this paper to evaluate a balance pointof design of cluster algorithm effect and to determine the optimum cluster result and itscorresponding tolerance value. Practices indicated that it is availability.The achievements researched in this paper were obtained by a series ofinnovative R&D work based on the TCM acupuncture information database system whichconstructed by related researchers in cooperating of Shanghai Research Institute of Acupunctureand Meridian and college of computer science and technology, DongHua University. Theexperiment results indicated that the expert system can meet core application requirements such asinquiring treatment approaches depending on a disease, inquiring diseases depending on medicaltreatment, knowledge updating and etc., have a better recall ratio and precision ratio and provideauxiliary decision basis for consulting of TCM acupuncture treatments, teaching and scienceresearch more well.