透過您的圖書館登入
IP:3.14.6.194
  • 學位論文

應用資料探勘技術於汽車業教育訓練課程分析

Applying the Data Mining Technology to Analysis of Job-training Programme in Motor Industry

指導教授 : 楊錦洲

摘要


企業為因應未來發展需要與解決現況面臨之問題,每年規劃各項教育訓練課程,由公司或單位主管指派學員參加,同時也開放其他員工報名選課。然而在經費有限之情況下,究竟該如何安排教育訓練課程?以及適當之選課對象應該為哪些學員?教育訓練後之學習成效是否符合公司要求?不同之訓練對象、選課組合與學習成效間有何相關?這些看似簡單問題,若能深入了解企業之訓練目的,並藉由統計分析工具,針對教育訓練資料庫進行探勘與分析,藉以找出學員屬性與課程間之特性,企業將可依此規劃更合宜之課程與安排更適當之學員受訓,達成提昇教育訓練效益之目的。 本研究係以資料探勘(Data Mining)工具,對某汽車公司人力資源訓練課程資料庫,先後採用交叉分析(Cross Table)、群集分析(Cluster Analysis)與決策樹演算法(Decision Tree)執行資料庫之探勘與分析,再依據分析結果提出教育訓練規劃作業流程改善建議,以供個案公司未來教育訓練課程開課時之開課策略參考,同時提升教育訓練課程規劃作業之效率。 本研究經進行資料探勘、分析與探討後,得到之結論有:根據過去教育訓練庫資料,以交叉分析找出學習成果與員工職位類別、員工部門類別及課程類別,其相互間具有顯著關聯性;以群集分析找出各職位類別與不同部門之員工熱門與較不熱門之教育訓練課程;同時找出同質性課程,未來可依據不同對象規劃一整套的訓練課程,讓員工有系統的參與課程,以提高教育訓練之效果,增加培訓合格率;以決策樹演算法探勘結果,個案公司2007年教育訓練投入成本中,未完訓者之成本約17.6%,其中又以高階主管之行銷部門之未完訓率最高,經此探討,有助於個案公司對學習成效與教育訓練規劃作業流程改善之重視,研究結果亦可供個案公司未來課程規劃之參考。

並列摘要


In order to meet the future requirement and solve the current problems, enterprises often plan to organize variety of Job-training for assigning employees and others. Nevertheless, how to arrange Job-training within limited budget? And who should be the group of students to take the course? Does the result of Job-training in accordance with the company requirement? What is the connection among training with different people, program module, and final outcome? All these questions might be easy to answer, but if the training purpose of a company can be fully understand, on the other hand, by using database mining and analysis will be able to find out the actual connection between students and training course. Therefore, the company can use the result to arrange more properly and achieve the effectiveness. This project uses the data mining tool to take a certain automotive company’s Job-training database analysis. We adopt the Cross Table、Cluster Analysis and Decision Tree to execute database investigation and analysis. According the results, we provided the suggestion for the improvement in Job-training process operation. It can become a reference for Job-training efficient lifting and Job-training strategy decision. After data mining、analysis and discussion, the conclusion is as following:By Cross Table, we can find out a apparent relativity in Job-training outcome with employee job rating 、department type and course assortment. By Cluster Analysis, we can pick up the relevant between job / dept. rating and course attraction. It would help us to arrange a series of suitable course for certain employee for Job-training efficient lifting. By Decision Tree, we find out the uncompleted course took 17.6﹪Job-training budget for this company case in 2007. Meanwhile, the high level manager in marketing division is key factor in all uncompleted course. By go through this research, it is helpful to make enterprise appreciate the Job-training efficient and the improvement for Job-training process operation. Meanwhile, this case research can be a reference for enterprise Job-training arrangement in future.

並列關鍵字

Cross Table Cluster Analysi Data Mining Job-training

參考文獻


楊琇媛(2003),「利用資料倉儲與資料探勘技術於招生策略與學生特質分析之研究」,私立中原大學資訊管理學系碩士學士論文
劉佳灝、饒瑞佶(2007),「學生選課組合與學習成效的資料探勘」,建國科大學報:管理類,第26卷,第2期,1-18
吳瓊治(2008),「如何運用ISO10015提升教育訓練品質」,品質月刊,第44卷,第3期,72-76
林文燦、柯美珠、王文杉(2007),「以資料探勘技術應用於教育訓練課程之規劃~以某A汽車公司為例」,品質學會第43屆年會年會暨第13屆全國品質研討會
李弘暉、吳瓊治(2007),「運用ISO10015轉換教育訓練投資位企業競爭優勢」,品質月刊,第43卷,第7期,78-83

延伸閱讀