摘要 隨著生物、資訊科技的迅速發展,研究主題多元化,典籍數量日增,驅動知識經濟時代的來臨,一向被認為傳統古典的本草典籍也被賦予新生命。典籍的整理、編纂、研究、實證與研究成果,皆必須進行數位化與資訊化。為提供各研究需求則必須建構完整的知識資料庫及資料探勘技術,且具備完整的搜尋功能,讓使用者很快的滿足其所需求的本草知識或資料。 中藥本草典籍中藥品的「主治(功效)」資料是中藥藥品得以流傳久遠的重要資訊,若能在中醫藥資訊化系統中提供正確無訛之屬性關鍵字詞,再輔以具迅速資料處理之資訊系統工具及網路之強大搜尋功能,必能建構更有效率的本草查詢資料庫、提升龐大本草藥品之資源應用。 本研究以「關鍵字擷選」之概念,將唐代「新修本草」、明代「證類備急本草」及宋代「本草綱目」三中藥本草典籍共4486筆藥物主治(功效)內文做為統計樣本,尋其「主治(功效)」之關鍵字詞,以字詞出現頻率及文件探勘的方法進行計算統計,以求關鍵字詞在中藥本草典籍重覆性出現頻率,並將結果輔以權重調整理論計算比較分析得知重要性排名。 希冀此研究結果對相關從業人員及研究人員能有所裨益、供予參考之價值,助益中藥本草典籍資料庫建置及查詢之相關研究,為中醫藥學之發展盡一份綿薄之力。 關鍵詞:中藥本草典籍、關鍵字、擷選、文件探勘
Abstract Era of knowledge-based economy has come and the ancient books of Chinese herbal medicine which have been regarded as traditional classics have been given new life along with the rapid development of biological and informational technology, the diversity of study, the increasing number of ancient books. To provide with the demands of the study, it has to build up an integrated science database and data survey technique with possess complete searching function. The user can rapid make use of the herb data or science with content. The data of “main cure” and “effect” of Chinese ancient herbal books is the important information that makes Chinese herbal spread age by age.If the key words of attributes can be correctly provided in the information system of Chinese herbal medicine, with the information system tools that can deal with data rapidly and strong function of web search, the data base with more efficiency for inquiry to Chinese herbal medicine can be conducted and the application of the resource of huge Chinese herbal medicine can be upgraded. This research bases on the concept of " key word capture selection", and takes three important Chinese ancient herbal books such as " mendation of herb " of Tang Dynasty, "Chinese herbal medicine" of Ming Dynasty, and "Chinese Herbal Medical Materials " of Song Dynasty, containing about 4486 articles of the mentioned three Chinese ancient herbal books to be used as statistical samples. Looking for the key word of "effect" and "main cure", it makes use of the way of the words and phrases appear frequency and document mining to carry on calculating statistics analysis to find out the repeated appear frequency of those key words in Chinese ancient herbal books, and to adjust theories with the assessment of the outcome to make comparison with and to analyze to compute and according to significance to put name in order. Hope the result of this research is helpful to the people who study and operate Chinese herbal medicine as reference. Neverthless, this research can be helpful to the conduction of the data base and inquiry of Chinese herbal medicine. Therefore, hope this research can be beneficiary to the development of Chinese herbal medicine. Key word: chinese ancient herbal book,key words, capture selection , documentary mining