以案例式推理及關聯規則建置問題診斷處理系統

Abstract

隨著企業對電腦和系統的依存度越來越高，早期使用者發現系統有問題，聯絡系統管理者處理的方式，已經難以滿足使用者需求。為了提供使用者穩定的資訊系統，系統管理者早已利用許多監控工具監控系統的健康狀態。然而，伴隨著資訊系統越來越多、越來越複雜，系統管理者監控資訊系統、處理被監控系統的異常項目也跟著越來越困難，系統管理者已不像以往能容易地、快速地提供使用者穩定的資訊系統。對於與資訊系統密不可分的企業來說，若恰好有一重要資訊系統的監控項目異常未被系統管理者即時處理，除了影響企業運作外，嚴重的話，可能會為企業帶來重大的損失。因此，本研究從如何提昇系統管理者管理資訊系統的角度出發，以資料探勘、案例式推理、關聯規則等方法學為基礎，實作一問題診斷的處理系統，運用案例式推理與關聯規則找出異常監控項目有效的解決方法給系統管理者，並透過關聯規則找出監控項目之間的關聯，進而提供相關知識與專家資訊，以期能協助系統管理者更有效的處理異常資訊系統。

As there is a great demand of using computers and systems for companies, the solution applied for those end-users to contact the system administrator directly while there is any problem occurred in the system could not meet those end-users’ requirements anymore. With the aim of providing end-users a more secure information system, the system administrator has already applied a variety of monitor appliances to examine the condition of the system. However, while the information system has increased significantly and become more and more complicated, it turns to become more difficult for administrators to manage, scrutinize those abnormal system problems easily and quickly, and provide users a safe system as it used to be. For those companies who are rely on these kinds of information system closely, if there is no monitor appliance can be employed to manage these situation in real time, it will have an impact on not only operating scheme, but also causing a great damage on those companies. Consequently, this study is aim to how to advance those information systems from administrators’ angle by implementing methodology based on data mining, case-based reasoning, and Association Rules. By employing the process flow of examining the problem, it can facilitate administrator to detect those anomalous monitor problems sufficiently based on applying Case-Based Reasoning and Association Rules. The implementation of Association Rule can ascertain the correlation among those monitor problems in order to provide some associated knowledge and professional guidance to administrator to manage the anomalous information system.