在先前的研究中，候選關鍵績效指標的數量不足以被分類至流程步驟之中，而分類的原因是因為關鍵績效指標可用來衡量企業流程的績效；除此之外，一個關鍵績效指標就產生一個多維度模型會使過多相似的維度模型被產生。在這次的研究裡，我們希望可以產生針對使用者需求數量的關鍵績效指標，並且藉由多維度表現指標的管理意義跟衡量流程的績效。最後，還希望能減少多維度模型的數量以簡化多維度模型的呈現。 此研究中，自動產生流程導向之關鍵績效指標的方法比較了企業資源規劃(ERP)資料庫之實體屬性和資料倉儲(DW)系統中既有之指標運算域的敘述相似度，並根據流程步驟來產生實體屬性跟運算域的連結，還透過預先定義好的公式結構去組合出候選關鍵績效指標。最後，我們建立候選關鍵績效指標之多維度模型並簡化多維度模型的結構，根據相同的流程步驟和相同的多維度模型，整合了候選關鍵績效指標的多維度模型。 我們的方法論提供了自動化的方法去產生更多的流程導向之候選關鍵績效指標和指標的多維度模型。首先，我們確切增加了候選關鍵績效指標的數量；接著還產生了以流程為導向之候選關鍵績效指標；最後，整合了候選關鍵績效指標的多維度模型根據相同的事實資料表、維度資料表、和流程步驟。 In previous research, the number of generated KPI candidates is not enough for classifying KPI candidates to step of the process because the KPI should be used for measuring the business processes. Otherwise, one KPI candidate has one dimensional model causes many similar dimensional model generated. In this research, we hope we can generate the user-specified number of KPI candidates which have dimensions to show the management meaning and measure the process. Finally, we decrease the number of the dimensional models to show the simplified dimensional model. In this research, the automatic step-oriented KPIs generating method compares the description similarity of Entity-attribute in ERP database with the operand of existing KPI in Data Warehouse (DW) system by text mining technique to make the linking of Entity-attribute and operand depend on the steps of the process. And it uses the defined formula structure to compose the KPI candidates. Finally, we construct the dimensional model of KPI candidates and simplify the dimensional model structure. We combine the dimensional models together depend on the same step of process and the same the dimensional model of KPI candidates. Our method provides an automatic approach to generate more step-oriented KPI candidates and its dimensional model. First, we increase the number of the KPI candidates. Second, we generate the step-oriented KPI candidates. Finally, combine the dimensional model of KPI candidate with the same fact table, the same dimension tables, and the same step of process.
