雲林縣政府於2003年開始陸續推動知識管理系統，提供文件儲存與檢索、電子表單、公告系統等功能。雖然資訊系統大部份都會留存使用行為紀錄，然而在雲林縣政府推動知識管理系統的過程中，使用者或相關系統的分析研究卻付之闕如，因此本研究整合Proxy紀錄檔與員工資料檔，以發掘知識，提供擬訂相關資訊系統發展政策。本研究的主要貢獻在於首次整合分析層級程序法(Analytic Hierarchy Process)及資料探勘(Data Mining)技術，將其應用在政府相關資訊系統的評估。本研究利用分析層級程序法以取得各系統之貢獻權重，並利用相關權重，結合資料探勘方法，以評估雲林縣政府內使用知識管理系統的員工。資料探勘所採用之演算法分別為K-means演算法及決策樹(Decision Tree)演算法，所分析之系統重要性及員工資料結果，將可提供雲林縣政府擬定新的資訊系統發展策略。

Since 2003, The Yunlin County Government starts to use a knowledge management system for document retrieval, electronic form and announcement systems etc. Most of the information systems can provide the information of user behavior by analyzing user's logs. However, while Yunlin County Government promotes the knowledge management system, the related analytical research of the date of users' background and users' logs are yet still lacking. Therefore, the research employs data mining techniques to analyze proxy server records and employee's data, and thereby to excavate the information of employee's behavior of using the system. As a result, managers of the Government can refer the results to make appropriate information system management policies for the employee in the Government. This research proposes a novel technique which simultaneously employs the Analytic Hierarchy Process and the Data Mining in order to specify the staff who frequently use the system in Yunlin County Government. Furthermore, we take advantage of the Analytic Hierarchy Process to find out the weighting of Information System, and combined the Data Mining with the purpose of revealing the user's behavior. Finally, the weighting exposes the significance of Information System, also, the mining results provide the knowledge of system's user behavior for manager to make strategies for Information System.