The method and application to construct experience recommendation platform of acupuncture ancient books based on data mining technology / 中国针灸

ABSTRACT

<p><b>OBJECTIVE</b>To construct a knowledge platform of acupuncture ancient books based on data mining technology, and to provide retrieval service for users.</p><p><b>METHODS</b>The Oracle 10 g database was applied and JAVA was selected as development language; based on the standard library and ancient books database established by manual entry, a variety of data mining technologies, including word segmentation, speech tagging, dependency analysis, rule extraction, similarity calculation, ambiguity analysis, supervised classification technology were applied to achieve text automatic extraction of ancient books; in the last, through association mining and decision analysis, the comprehensive and intelligent analysis of disease and symptom, meridians, acupoints, rules of acupuncture and moxibustion in acupuncture ancient books were realized, and retrieval service was provided for users through structure of browser/server (B/S).</p><p><b>RESULTS</b>The platform realized full-text retrieval, word frequency analysis and association analysis; when diseases or acupoints were searched, the frequencies of meridian, acupoints (diseases) and techniques were presented from high to low, meanwhile the support degree and confidence coefficient between disease and acupoints (special acupoint), acupoints and acupoints in prescription, disease or acupoints and technique were presented.</p><p><b>CONCLUSIONS</b>The experience platform of acupuncture ancient books based on data mining technology could be used as a reference for selection of disease, meridian and acupoint in clinical treatment and education of acupuncture and moxibustion.</p>