鑑於近幾十年來氣象逐漸異常，各行各業在決定方針與策略時，開始顧慮氣象因素，進而發展出氣象經濟與氣象醫學等等與氣象有關之事物。本研究利用資料倉儲之萃取、轉換與載入之概念，以癌症病患中醫證候與氣象因子為例，將此兩種資料進行整合，建構出中醫證候與氣象因子資料方體，從資料方體取出不同時間維度資料進行敘述性統計，再從資料方體取出資料探勘所需之資料，使用RapidMiner資料探勘軟體進行決策樹分析，找出適合不同型態之輸入變數模型，最後上述兩種方式，找出在12、1月與10、11月時，平均氣壓最影響數值型態與最低氣溫最影響類別型態，與其他氣象因子在不同規則下皆會影響肝火上炎證候。 In view of the gradual abnormality of meteorology in recent decades, various industries began to consider meteorological factors when deciding policies and strategies, and then developed meteorological economy and meteorological medicine and other meteorological-related things. In this study, the concepts of extraction, transform and loading of data warehouses are used to take the TCM syndromes of cancer patients and meteorological factors as examples, and the two data are integrated to construct a cube of TCM syndromes and meteorological factors. From the data cube extracts data from different time dimensions for descriptive statistics, and then extracts the data required for data mining from the data cube, and uses RapidMiner data mining software for decision tree analysis to find input variable models suitable for different types. Finally, the above two methods find out that in December, January, October, and November, the average air pressure most affects the numerical type and the minimum temperature most affects the nominal type, and other meteorological factors under different rules will affect syndromes of Flaring-up of Liver Fire.
