- 學位論文
利用事件軌跡與資料庫軌跡進行商業流程分析
Business Process Analysis based on Event Logs and Database Logs
摘要
目前企業導入各類資訊系統，其目的是為了提升企業的執行效率與競爭力，但往往作業流程與系統運作無法如預期般的密切配合，歸納常見的問題不外乎系統與作業流程規劃不夠嚴謹，或例外狀況處置不夠完善等，要如何改善現行已知的缺失並全面檢視發現未知的問題，商業流程再造將是一個改善缺失與提升效率最好的方法之一，透過商業分析可以幫助企業管理者與審計人員發現例外的流程，大多數商業流程分析的研究在都是利用資訊系統內的日誌分析，利用α演算法找出商業流程並由派翠網（Petri Net）來呈現，或是分析資訊系統內的資料表的主鍵與外來鍵的關係與日誌來分析商業流程，這些研究的先決條件為資訊系統日誌紀錄實例(Instance)的因果關係，但一般而言資訊系統的日誌記錄通常都是缺少這些資訊，故本研究的目的設法找出實例的因果關係，藉此可以分析商業流程，且配合資料庫的稽核日誌分析商業流程中每個活動所使用到的資料表，故此研究提供一套邏輯方法來分析企業內各資訊系統之資訊或外部不同資訊之紀錄與追蹤，藉此追蹤資訊系統內部之流程問題，並整合事件軌跡(Event Log)與資料庫軌跡(Database Log)進行商業流程分析，利用Event Log與Database Log，發展一套分析作業邏輯關連因果法則之方法，以檢視流程活動是否有違常理，並提供未來企業進行流程重新設計的基礎參考資料。
並列摘要
Enterprises introduce various information systems to improve execution efficiency and competitiveness. However, they may usually encounter incompatibility of the systems’ operations with existing operating processes. Common causes of this problem include insufficient planning of the systems, operating processes and inadequate handling of exceptions. How to improve known problems and excavate unknown ones is an important issue. Business process reengineering is one of the best solutions. Business analysis helps corporate managers and auditors discover exceptional processes. In business analysis, most studies rely on the logs of information systems. They use α-algorithm to explore business processes and Petri Net to represent them or analyze the relationship between primary keys and foreign keys and logs to analyze business processes. The analysis of these studies is, however, built upon the availability of causal relationships among instances in information system logs. Generally, causal relationships among instances are absent in most information system logs. This study managed to find the causal relationships among instances and attempted to use them along with audit logs to analyze tables used for each activity in the business process. Hence, this study provided a logical method for analyzing logs and records of all information systems, both internal and external. By integrating event logs and database logs for business process analysis, this study developed an operation association analysis method that could examine if any activity in the process is in violation of common norms and also serve as a basis for business process reengineering.