中醫診斷分為望、聞、問、切(通稱四診)，透過四診得到病患所有的資訊，醫生藉由這些資訊去分析、診斷及處方用藥，此過程就是辯證。因此，症狀是疾病診斷的重要依據[1] ，然中醫發展悠久，諸多因素導致症狀描述不盡相同，如能將中醫症狀描述做標準化處理，對中醫辯證以及科學化分析有很大的助益[2] 。而且經由對於標準化症狀反復的探討及修正，可以發現症狀與症狀的組合規律，形成中醫的證候及病因病機的主要內容。本研究使用《中醫證候學》[3] 的症狀描述做為標準化對象。書中對於證候、證象以及症狀的關係有完整的架構呈現。本實驗設計一套症狀標準化流程，且提出驗證標準化症狀的方式，透過症狀分群去分析證象的關聯性，觀察分群結果，修正標準化錯誤的症狀，重新分群直到結果符合書中的證象分群。以便為往後的專家辨證系統奠下最重要的基礎。

Syndrome pattern is one of core theories of Traditional Chinese Medicine. Symptom representthe most important sources of the information required for pattern identification. Symptoms areoften due to geographical, climate, customs and physicians have different background narrative.These annoy doctors and researchers.Therefore, we proposed the concept of “Symptom Standardization”. There exists 3516sentences in Theory of Traditional Chinese Medicine Syndrome-complexes which describessymptoms and we would like to present them in the following manners:1. One symptom unit only express one connotation.2. One symptom has its own classification , properties and its property values.3. Quantify a symptom description of the attribute4. Unify a symptom description of the attribute.5. Sentences with same meaning must standardize to the same result.Symptom Standardization is used to produce traditional Chinese medicine symptomsQuestionnaire to facilitate the collection of disease symptoms, and also used to analysis withalgorithm to verify the theory of traditional Chinese medicine. This study will describe howsymptoms standardization and its procedures, clustering the standardization symptom usingstatistical probability, through the clustering results verify standardization symptom.