Data Mining in Traditional Chinese Medicine Research Academic Schools
Author: Tan Yong, Lu Ai-ping, car read-chung, Cheng Lü , Cai Nian-
Ning, Zhang, the Jiang Chunyan
[Abstract] schools in the academic study of traditional Chinese medicine there are a lot
of data, using data mining techniques can get hidden in these data for useful information. By citing the example of Zhao Bingnan academic research schools, briefly explained the use of data mining of academic schools of Chinese medicine practitioners
and advanced the feasibility study and proposed the implementation of data mining steps and specific tools, algorithm.
[Key words] Chinese Academic Schools Data Mining Psoriasis
Since ancient times, Chinese academic schools everywhere, not only contending
with each other, they interpenetrate each other. To promote their development and research in basic theories of Chinese medicine has been gradually expanded and deepened to promote the clinical practice of Chinese medicine has been a gradual
progress in technology and enhanced. Can be said that the academic schools of Chinese medicine development and research to promote the development of Chinese medicine .
The premise of the academic study of traditional Chinese medicine schools is to
divide and to determine their scientific research content. With the scientific and technological advances and the growing Chinese extensive academic research, academic schools of Chinese medicine into continuous improvement methods to
improve. Xu Jiang-Yan et al  proposed classification model of the core of both the Han-na vertical transmission of academic thinking, but also compatible with the horizontal infiltration of academic thought, there is a certain degree of progressiveness,
it is worth learning. We are in the 'traditional Chinese medicine dermatology Zhao Bingnan Academic Schools and transmission studies' subjects, from the following two aspects of school division and determine the research content: First, with innovative
academic ideas as the core: Based on Chinese medicine dermatology Zhao Bingnan Academic Schools Academic Origin basic theory, characteristics, experience, technology and methods of systematic research on Chinese medicine dermatology Zhao Bingnan academic schools to establish academic ideology; secondly, in order to open the medical home for the nuclear group and its writing quality: whether Zhao Bingnan portal modeled after the income, or the Gambling public house is said, but also whether published papers, on the meeting written communication materials, scientific research, inventions, patents, or oral testimony, as long as the academic schools of Chinese medicine dermatology Zhao Bingnan relevant content, all include in their nuclear quality of being.
In response to these research content, in accordance with previous research methods, can people with the disease, to prescriptions or way of thinking leads to literature for the collation, analysis and conclusion, or clinical epidemiology methods of
clinical observation and conclusion [3 ~ 7]. These methods can be summed up in a certain extent, the law to guide clinical practice, but there are the following shortcomings: ? the experience of academic ideas and clinics associated with the lack
of systematic research and comparative study of organic; ? difficult to reflect the
course of the disease clinics, multi-dimensional space-time TCM and non-linear
characteristics; ? consistent with traditional Chinese medicine characterized by the lack of multi-factor information-processing technologies, research results might lead to confusion and error; ? difficult to truly reflect the academic schools of Chinese
medicine meaning and essence. In view of the problems of previous studies, we have introduced the data mining technology to the academic schools of traditional Chinese medicine research.
1 Data Mining and Its Application in the field of medicine
1.1 Overview of Data Mining Data Mining (DM: DataMining) is from a large number of, incomplete, and vague, random data, extracting implicit in the work, that people do not know in advance, but is a potentially useful information and knowledge The process . It is a cross-discipline, brings together the database, artificial
intelligence, statistics, visualization, parallel computing in different disciplines and fields. In which artificial intelligence to mimic human non-linear, incomplete, imprecise
and uncertain information in intelligent processing, has a good fault tolerance, robustness and high-precision integrated technical advantages such as [9,10]. Artificial intelligence methods include artificial neural networks, fuzzy systems, evolutionary computation, rough set theory, support vector machine (SVM) and so on. Excavation for different objectives, often several ways to integrate applications to exploit their own technological superiority. In addition, artificial intelligence and other data mining methods such as cluster analysis, decision trees, association rules, and optimize applications, it can bring the data mining technology to solve complex problems.
1.2 Data Mining in the field of medicine the application of data mining technology for medical data polymorphism, no integrity, timeliness, and redundancy features of
the implementation of sound data processing and knowledge extraction . It is used mainly in the field of medicine in the following aspects [12 ~ 16]:
1.2.1 drug development research to a new, built on the full use of several thousand accumulated wealth of experience and modern science and technology, the basis of information provided by the model studied.
1.2.2 Compatibility of compound and the side permits the corresponding data mining technology can be to a certain extent, the discovery and understanding of the
clinical conditions are associated with re-enact Group-West relations, compound drugs,
compatibility relations, Medicine Smell the interaction between the relationships.
1.2.3 Research on the medical information in the ancient language and text-based
Chinese medicine theory and practice of structured analysis is the study of medical information important content, including by some of the content can be text data mining to achieved.
1.2.4 Expert System of Traditional Chinese Medicine research on Chinese
medicine experts 'can only realize, unspeakable' grasp of the law of diagnosis and treatment of disease, qualitative description, fuzzy concepts, using data mining technology for intelligent analysis of multi-dimensional, to a certain extent, they can be
understandable rules or models to be expressed, which have greatly enriched the expert system knowledge base.
1.2.5 Traditional Chinese Medicine literature data mining technology in literature can improve the level of literature to improve the efficiency of the use of literature, and thus can improve the efficiency of scientific research work of Chinese medicine.
2 Data Mining in Traditional Chinese Medicine Research Academic Schools
With 'traditional Chinese medicine dermatology Zhao Bingnan Academic Schools and inheritance studies' in the psoriasis data mining, for example, a brief description of the steps in mining and mining tools and algorithms.
2.1 The steps in the implementation of data mining
According to the world's first universal model for data mining industry standards (CRISP-DM), psoriasis data mining process can be divided into six stages.
2.1.1 understand the problem clear and detailed research goals.
2.1.2 understand the data collected are familiar with the type of data related to psoriasis and the presence of form, the preliminary data based on the quality of identification, and establishment of databases.
2.1.3 The preparation of data related to the data in psoriasis there is a large
number of incomplete, redundant and inconsistent data. When conducting data mining, we must first be processed. Process including data extraction, cleansing, transformation and loading.
2.1.4 the establishment of the model addressed the research objectives, using the known data and knowledge of the establishment of analytical models, and effective application of the model to the unknown data or similar cases, tested and modified
model, and so repeatedly to get the optimal model.
2.1.5 Program assessment in the model before the final implementation, a thorough assessment of the model, and then review the steps of the model constructed in order to determine the model that can really achieve the intended goal of the
2.1.6 The implementation of the program based on the above five aspects of the work, using the corresponding data mining tools and algorithms repeatedly, adjusted several times, and constantly revising and improving the data mining. Reposted
elsewhere in the paper for free download http://
2.2 Data Mining Tools and Algorithms
2.2.1 Psoriasis Psoriasis TCM Syndrome normative research in the database, there are many uncertain, incomplete, inaccurate information about the dialectic, the first to
use rough set them about Jane, excluding vague and ambiguous information, reservations symptoms with the excavation of important attributes. On this basis, using the method of mining association rules permit the distribution of cards than the
composition, which are static correlation. Using FP-growth or weighted Apriori
algorithm to find frequent item sets to determine the association rules. In the correlation analysis based on similarity clustering, by a given threshold, the
composition of the symptoms and the symptoms of the contribution rate, to be symptom-owned permit clustering. Using a genetic algorithm search and K-means
combination of local optimization, according to a recent genetic cross-matching
operator, in the course of the continuous generation of new cross-individual, to ensure
the diversity of groups, reducing the K-means algorithm premature phenomena and
solve the global optimization excellent question. K-means clustering can be found
locally specific symptoms, point algorithm used to isolate them special treatment. After clustering all of the symptoms, will be N a cluster. Because the data mining results are not subjective factors, so these clustering need Picco experts, based on traditional
Chinese medicine clinical practice be renamed. In the syndrome classification based on the use of time sequential pattern mining method to analyze the evolution of course the basic trend of each card.
2.2.2 Western psoriasis disease, cards, disease and objective examination of the
relevance of indicators of research in psoriasis database, for Western relevant symptoms, signs of the feature extraction method using association rules to establish the symptoms, signs and objective examination of indicators the relevance model, and
then into the TCM classification rule induction in order to establish the correlation between their networks.
2.2.3 psoriasis card, disease and objective examination and treatment of herbs and different indicators of correspondence study in the treatment of psoriasis in the
database, using data mining classification to a single taste of Chinese medicine treatment of psoriasis to classify projections in order to improve its potency. And use decision trees and association rules analysis of drug efficacy classification with their potency characteristics of the relationship between the association, using rough set to simplify drug potency characteristics. Treatment of psoriasis dose of compound Medicine Smell and varied, the use of decision trees, association rules, and attribute-
oriented induction method of analysis of different compatibility levels (single herb medicine, drug efficacy classification, etc.) Medicine Smell the relationship between compatibility and use support vector machines and Bayesian networks to classify in the compound, based on the common understanding of the laws of the compound, using rough sets to realize a simplified and characteristics of compound extraction. Chinese medicine treatment of psoriasis have different treatments, using clustering and related algorithms to determine their indications. Psoriasis of the card, disease and objective examination and treatment of herbs and different indicators of the corresponding relationship between therapy mainly uses genetic algorithm simulation study on their corresponding non-linear relationship between the global nature of search, to find out the optimal state of correspondence.
2.2.4 Clinical Evaluation of Traditional Chinese Medicine Psoriasis research in
psoriasis database, the use of information extraction technology was able to reflect the clinical efficacy of traditional Chinese medicine effective indicators (mainly refers to the symptoms and objective examination of indicators), and the use of clustering and
time series model Mining can objectively reflect the clinical efficacy of medical
interventions for psoriasis evaluation indicators and methods, in particular, must be clear of psoriasis symptoms throughout the course of the data in the evaluation of the
efficacy of the contribution.
The above four studies have focused on are also closely linked, they are involved in a number of technical aspects of data mining tools and algorithms. While these tools and algorithms independent of each other, but they have some complementary, and only in time and space on the optimal combination to get the satisfaction of working closely with the mining results.
Data Mining for the Study of Schools of Traditional Chinese Medicine provides a
powerful tool. Through which to expand and deepen the connotation of Chinese academic research schools, on the one hand the performance of the theory of sublimation in the promotion and innovation, on the other, in favor of modern science-
based disease, the discovery of the laws governing permits, more importantly, is seen in the disease clinics help introduce new technologies, and advancing with the times. These three aspects of traditional Chinese medicine effective in improving the academic
research level and quality of schools, and promoted the development of academic medicine.
 Hu Bin. An Essay on the genre of academic medicine [J]. Chinese Journal of History, 2004,22 (4): 1.
 Xu Jiang-Yan, Xie Yanggu, Lu Zhaolin. TCM Academic Schools interpretation [J]. Journal of Beijing University of Chinese Medicine, 2003, 26 (3): 15.
 Qiu Ji, Zhu Jianping. Zhu Liangchun Khan syndrome differentiation and treatment of sleep medicine experience, Selection and Analysis - Famous TCM
professor Zhu Liangchun clinical experience (37) [J]. Liaoning Journal of Traditional Chinese Medicine, 2003,30 (1): 14.
 Hao Cheng-yan, Li Zhenbin. Dynasties famous doctors of the drug treatment of abdominal pain in pregnancy analysis [J]. Chinese Medical Journal, 2002,17 (12): 723.
 Yan Jiang. Famous doctors finishing of Prescription drug laws reflect [J]. Journal of Beijing University of Chinese Medicine, 2003,26 (1): 15.
 LIU Yan-jiao. TCM clinical study to explore ways of thinking [J]. China Academy of Traditional Chinese Medicine Hospital reported that, 2003:7.
 WANG Ying-hui, Kang Yang, Yan Ying-jie, et al. Based on information and
data mining techniques were the old idea of TCM clinical experience study [J]. The
world of science and technology - in the modernization of Chinese medicine, 2005,7 (1): 98.
 FanM, MengXF. Datamining: conceptsandtechniques [J]. ChinaMachine-
Press: 2001: 1-322 [Fan, Xiao-Feng Meng translation. Data Mining: Concepts and
Techniques. Beijing: Mechanical Industry Press, 2001:1.]
 Coulter DM, Bate A, M eyboom RH, et al. Antipsychotic drugs and heart muscle disorder in international pharm acovigilance. Datamining study [J]. BMJ,
2001,322 (7296): 1207.
 Ohrn A, Row land T. Rough sets: a knowledge discovery technique formultifactorial medical outcome [J]. Am J Phys Med Rehabil, 2000,79 (1): 100.
 ZHU Ling-yun, WU Bao-ming, Cao Chang-xiu. Medical data mining
techniques, methods and applications [J]. Journal of Biomedical Engineering, 2003,20 (3): 559.
 Yao Mei, Ai Road, Yuan Yue-Mei, et al. Diabetes compound compatibility law
of association rules [J]. Journal of Beijing University of Chinese Medicine, 2002,25 (6): 48.
 Gu Zheng, Gu. Information Extraction from Chinese medicine research in the application. Medical information, 2005,18 (9): 45.
 Yang Chunhua, Li Chunhua, Wang Gui-zhi, et al. Data mining technology in
drug research and discovery application [J]. Medicine Herald, 2005,24 (12): 1143.
 Qin Zhang, Liu Ping, Zhang WT. Data mining technology in the study of TCM applications. Shanghai Journal of Traditional Chinese Medicine, 2006,40 (3): 3.
 Xie included. Data Mining in the literature of Chinese medicine [J]. In the medical information, 2005,22 (6): 5. Reposted elsewhere on the Download Center
http:// free papers