工作流程在時程上的管理上是一個很重要的議題，工作時程中異常的延遲時間將導致企業營運不彰，使得企業無法發揮最大的績效。因此本研究提出挖掘特例工作時程的演算法，根據過去工作時程記錄找出工作流程上異常延遲時間的活動，作為日後調整工作流程的參考依據。在管理上的意涵能輔助企業主管或顧問找出企業流程當中各個活動延遲的可能，透過改善以利日後有效控制特例的情況，另一方面則是可以方便管理者找出工作流程中時間異常的癥結點。 本研究中以活動為基礎，提出目前工作流程延遲時間最完整的可量化測量種類探討，而在演算法的設計部分則使用三維陣列存放方式，讓延遲時間計算上更有效率，並透過真實的資料所得到的實驗結果，提供管理者改善的方向。 The concept of workflow is an critical issue in time management. Irrational delay time not only causes a firm’s inefficient operations but also hinder a firm’s ability to optimize performance. This research provides an algorithm which based on workflow’s abnormal delay time. By finding out company’s workflow outlier and take it as a reference for adjusting the whole workflow, the algorithm presented in this research helps managers and consultants recognize all possible delay types of workflow’s activities as well as figure out the crucial activity of irregular instances of workflow easily. Building on the activity-based method, we explore possible types of workflow delay time and further propose complete uantifiable measurements. Three-dimension data store on struct is applied in the algorithm to compute time delay effectively. Besides, empirical data are drawn to validate our model as to provide practitioners additional instructions to improve operational performance.
