校務研究在近幾年已經成為許多學校的主要研究方向，目的是為了提 高學校本身的績效。雖然校務研究已經執行多年，對於小範圍議題之分析 已經很普遍，但對整體高教的探索性研究還尚未形成一個共識，以至於議 題的研究選擇重複性很高。其中造成議題重複性很高的原因為資料的缺 乏，在有限的欄位下，能探討的議題就極其有限，因此，為了達成廣泛的 探索性研究，需要大量的資料欄位來輔助，才能逐步擴大議題來分析，達 到廣泛探索之目的。 本研究模擬資料欄位擴增的情形，開發漸進式演算法應用於欄位擴增 的資料來降低時間成本，同時針對新加入的欄位，挖掘新舊欄位之間的關 聯，試圖尋找出新的欄位是否對於舊欄位之間的關聯產生影響，以及新舊 欄位尚未交互前無法挖掘出的關聯。藉由廣泛的探索資料的關聯來提供決 策者分析，達成決策支援。 ;Institutional research has become a major research direction for many schools to improve the school’s performance in recent years. Although Institutional research has been conducted for many years, analysis of small-scale issues is already common, but the exploratory research on higher education as a whole has not yet reached a consensus, so that the research selection of issues is highly repetitive. One of the reasons for this issues of repetition is the lack of data, and the limited number of attributes that can be explored with limited fields makes it extremely limited. Therefore, in order to achieve extensive exploratory research, a large number of data attributes are needed to assist in expanding the scope of analysis and achieve the goal of extensive exploration. This paper simulates the situation of expanding data attributes, develops algorithms applied to data attributes expansion to reduce time costs. In addition, for these expanding data attributes, trying to identify whether they have an impact on the correlations between old attributes, as well as correlations that cannot be identified before the interaction between expanding attributes and old attributes. Through extensive exploration of data correlations, decision-makers can be provided with analysis and achieve decision support.
