Knowledge Discovery and Data Mining on Clinical Case Reports of Traditional Chinese Medicine
Objective: clinical case reports from famous TCM physician are important recourses for study and learning. Tens of thousands of medical recorders which were saved by texts and data-bases have mass study variables and are typical multi-dimension data. We reveal characteristic clinical thinking mode and establish a novel TCM theory by mining and knowledge discovery of huge information from medical recorders from famous TCM physician. Methods will be screened and compared whether data mining mode and knowledge discovery are feasible for case reports. Methods: CRISP-DM international standard process was introduced, and computer information technology, database and statistics software were used to scrub and mine the data. Many methods of data mining like generalized dictionary model suitable for case data were introduced. Result: some useful conclusion was obtained and disclosed the features of TCM pathogenesis which is to summarize complicated conditions with simple rules. A novel TCM theory was created about Pathogenic element (PE) and Pathogenic unit (PU) which will direct a new study flied. At the same time some key formula of compatibility of ingredients were captured successfully in recorders. Conclusion: methods of sorting and clearing up complicated data are feasible, which ensure data to be true and avoid information loss. Generalized dictionary model can extracted rule of treatment determination based on syndrome differentiation and find new knowledge.
TCM case report data mining knowledge discovery
LI Guochun WU Mianhua YE Fang
Center of Chinese Medicine Statistics, Nanjing University of Chinese Medicine, PR.China, 210029 First Clinical College, Nanjing University of Chinese Medicine, P.R.China, 210029
国际会议
威海
英文
298-301
2010-07-24（万方平台首次上网日期，不代表论文的发表时间）