不同顆粒度流程一致化之研究; Unifying multi-level business process discovered by Heuristic miner Algorithm

NCUIR > School of Management at National Central University > Graduate Institute of Industrial Management > Electronic Thesis & Dissertation > Item 987654321/59939

Please use this identifier to cite or link to this item: http://ir.lib.ncu.edu.tw/handle/987654321/59939

Title:	不同顆粒度流程一致化之研究;Unifying multi-level business process discovered by Heuristic miner Algorithm
Authors:	陳弘皓;Chen,Hung-Hao
Contributors:	工業管理研究所
Keywords:	流程探勘;啟發式演算法;一致性檢定;石材加工;斐氏圖;Data mining;Heuristic miner algorithm;Conformance checking
Date:	2013-06-24
Issue Date:	2013-07-10 11:49:40 (UTC+8)
Publisher:	國立中央大學
Abstract:	近年來資訊系統的發達，整合資料庫資訊來提高可閱讀性是必要之路，所以許多企業都會將流程探勘技術運用在管理中，而企業流程管理與企業流程一致性在一家企業內部是相當重要的。當企業發展到一定規模的時候，管理企業就需要有系統的流程規劃，否則將會影響企業的發展，而近年來各種企業也越來越重視企業內部的控管，當實際預測的企業流程與預期的模型不一致，就表示這中間就可能有舞弊、錯誤的風險，亦顯示可能存有內部控制的缺失。當在檢視企業內部中事件日誌的流程記錄時，企業流程可以利用事件日誌回推，而因為事件日誌有不同的顆粒度，所以事件日誌中流程的顆粒度也有所不同，我們可依事件日誌的流程用grain 1到grain n來表示，而grain越低就代表流程顆粒度越小。當用事件日誌回推企業流程的時候，會發現從同一個資料庫顆粒度不同grain的流程記錄，回推出來的企業流程其中可能有所差異。而本研究演算法是要解決，當遇到前面所提到由同一個資料庫卻因為流程的顆粒度不同，發生回推出企業流程也不相同時，可以經由調整啟發式演算法(Heuristic miner)中所提到的相依關係係數(dependency threshold)來自動解決一致性差異的問題，而本研究利用石頭加工產業的流程事件日誌來進行案例驗證，在比較fine以及coarse兩種不同顆粒的企業流程一致性下，可以設定其中一個grain的Process Model相依門檻值以及一致比率後，系統自動給定另一個grain的Process Model相依門檻值，此種方式就可以讓企業在源自同一事件日誌卻因不同grain所推導出的企業流程中，較容易的找出對應的一致企業流程。 In recent years, many companies use the process mining techniques in the management. The companies will use the process mining techniques to derive the business model. When the Business model and the real process do not match, it means that there may be a risk inside the companies and it also shows the lack of internal controls that may exist. In the past research, conformance checking focused on the difference between event logs and business model. But in this research, when checking the process in the event logs, there are different grains of process in the same event logs. It can represent the process from grain 1 to grain n according to the event logs, and the lower of grains means the finer of process. When using the process in event logs to mine the business model, it shows that using different grains of process from same event logs will mine out different business model. The purpose in this research is solving the problem that different grains of process from the same event logs will mine the different business model by using the dependency threshold of Heuristic Miner Algorithm. This research use the event logs from a stone processing industry for case verification. By setting two different grains of process, the business model of fine and coarse to compare the consistency. If giving the dependency threshold and consistent ratio from one of business model, the system will automatically give another grain of the business model dependency threshold.
Appears in Collections:	[Graduate Institute of Industrial Management] Electronic Thesis & Dissertation

Files in This Item:

File	Description	Size	Format
index.html		0Kb	HTML	715	View/Open

社群 sharing

Loading...