以MDG為基礎之多媒體教材檢測工具之設計與實作

Abstract

數位學習日漸蓬勃發展，因此數位多媒體教材的品質也就日趨受到重視。目前有不少機構提出了多媒體教材的品質認證，如臺灣的數位學習品質服務中心。認證的方式主要是以教材成品使用檢核表 (Check lists) 對品質項目一一做確認。這種對「後期品質」認證方式最大的缺點就是當此時發現品質問題，要更正所花的成本會比在教材設計階段就發現問題而更正要來的高。除了用品質認證的方式之外，也有學者提出針對多媒體教材生命週期的品質確保模型。其中關於品質確保的方法為在教材製作的各個階段中，使用Perspective based Inspections、Testing、Prototyping、形成性評量以及總結性評量。這些方法大多數都是教材製作參與人員使用檢核表互相檢查，缺少可以自動化檢查教材的工具。

基於上述，本研究希望設計並實作一個多媒體教材品質檢測工具，方法是使用一多媒體品質管制模型─MDG (MLCs Directed Graph)，搭配現有的品質檢測演算法，期望能在教材製作中，能自動化檢測出教材之中的一些可能的品質不良情形，使問題能夠在教材製作的早期階段就被排除，進而達到教材品質提升的目的。

E-learning is becoming a blossomed method for learning, teaching and training, and assuring e-learning quality is becoming an important issue worldwide. Some organizations have proposed e-learning courseware quality assurance so far, for example, the eLQSC (e-Learning Quality Service Center) in Taiwan. The major quality assurance method is using quality checklists to evaluate e-learning courseware. One of the drawbacks of the “late” quality assurance is that solving problems in courseware during the late phase of courseware development life-cycle is more costly than solving them during the early phase such as design phase. In addition, a number of multimedia courseware development model through life-cycle have been proposed. They use perspective based inspection, testing, prototyping, formative and summative evaluation to evaluate a courseware quality, and lacking an auto-testing tool to testing problem of a courseware automatically.

According to the above reasons, this study wants to design and implement a multimedia courseware testing tool using the courseware quality control model MDG (MLCs Directed Graph) and existing quality testing algorithms, expecting the quality issues can be auto-tested during the early phase in the development life-cycle and improve the quality of a courseware.