資料探勘之技術已成為近年來企業及訓練單位取得重要決策資料的工具之一,而資料探勘就是以統計學和電腦科學為基礎,所發展出來能快速分析資料的方法,其探勘工具也因資訊科技之進步讓使用者能有多樣的選擇。 本研究依據職訓局就訓練品質評核系統(Taiwan TrainQuali System,TTQS)[9]之需求,在其評核計分表中,以標竿單位其成功經驗可做為其他業界之典範[11];但對於過去受評單位皆無法從中取得關鍵指標足以提升其訓練品質系統,透過資料探勘之群集分析特性可快速得知及建構成功關鍵指標並可解決此問題。 本研究以群集分析法作為資料探勘之工具,以辦理教育訓練單位為主要研究對象,針對TTQS各項指標分數進行資料分析,以群集分析方式產出關鍵指標,再進行資料比對與分析,產生資料回饋,此一模型可提供相關研究數據可供日後訓練單位執行及研究參考。
In recent years, data mining technology has become one of the important tools to obtain information for decision-making by enterprises and training institutions. Data mining is based on statistics and computer science to develop methods that can quickly analyze data. As a result of advances in information technology users may be able to have a variety of choices in mining tools . This study is made in accordance with Taiwan TrainQuali System (TTQS) developed by Bureau of Employment and Vocational Training , After the benchmark units have imported their data into TTQS, the successful results can be used as a model for other industries. However, the disadvantage is that previous units being badly scored were not able to obtain key success indicators to upgrade training quality. Therefore, this study will apply a cluster analysis approach to resolve the above problem, through database analysis, and develop a set of key success indicators to be used later on by other training units as a reference for their training quality analysis. This study applies cluster analysis approach as a tool for data mining with the training units as main object of study. The various index scores of TTQS are used for data analysis, and clustering methods will produce the output of key success indicators. Next, data are compared and analyzed, and finally information feedback is produced. Such a model can provide relevant training units with research data for their references.