戴建耘林培倫2019-09-042006-6-302019-09-042006http://etds.lib.ntnu.edu.tw/cgi-bin/gs32/gsweb.cgi?o=dstdcdr&s=id=%22G0069270029%22.&%22.id.&http://rportal.lib.ntnu.edu.tw:80/handle/20.500.12235/98478本研究主要是針對教育部大學課程資源網之資料倉儲中存放的課程資料進行研究。以「線上分析處理」技術為概念，並配合各個不同領域專家學者的研究需求建構客製化模板分析工具。最後以大學院校89~93學年度傳輸網路光纖科目課程資料為例進行分析及驗證。 資料倉儲內課程資料經蒐集、轉換、清洗等流程後，建立資料超市，再將課程資料送入本研究所建構之模板，由「各主題分析圖表之自動產生引擎」處理後產生具多維度之圖表。最後藉由各領域專家學者的觀點來觀察圖形及表格，依照其專家領域經驗進行判斷，從而了解存放在資料倉儲中資料之現況和趨勢所在，並可從其中找出隱藏、未知的資訊，未來針對大學課程資料倉儲內儲之課程資料進行研究時，有了一個適合的分析工具可使用。 　　本研究所建構之客製化模板分析工具有以下幾項特點： 一．可依各個不同研究領域之專家學者需求來建構分析模板。 二．經由客製化模板分析工具，可對資料倉儲中所存資料進行多維度分析。 三．分析模板可產生各式各樣簡單易懂的報表、圖表，提供操作者使用。 四. 各領域專家學者可在圖形、表格中，依不同角度和觀點發現隱藏、未知、獨特的資訊與知識。This research focuses on the curriculum data stored in the data warehouse of the Ministry of Education’s College Curriculum Resource Network. Based on the On-Line Analysis Processing technology, customized analysis models are built to meet the research needs of experts and scholars of various fields. Curriculum data for Optical Fiber Transmission Network for years 2000 ~ 2004 was analyzed and verified as an example for this research. The curriculum data in the data warehouse was collected, transformed, filtered, and was used to build a data supermarket. Then the curriculum data was fed to the customized models developed in this research. Multi-dimensional graphs and tables were created using “Topic-Specific Analysis Graph and Table Automatic Generator.” Inspecting the graphs and tables from expert’s point of view and judging them based on expert’s knowledge and experience lead to understanding of the current status and trends of curriculum data stored in the College Curriculum Resource Network, and discovering of the information it implies. It also provided an appropriate analysis tool for future research on data stored in the College Curriculum Resource Network. The customized analytic template developed in this research have the following features: 1.The analytic template can be developed according to the needs of different experts and scholars of various research fields. 2.Multi-dimensional analysis can be performed on data stored in the data warehouse through customized analysis models. 3.Analysis models can generate varieties of user-friendly reports, graphs, and tables for users. 4.Experts and scholars of different fields can discover hidden, innovative, and unique information and knowledge from different angles and point-of-views.資料倉儲資料挖掘線上分析處理客製化Data WarehouseData MiningOLAPCustomization