多功能數位運動知識庫的塑模－以棒球為例

中文摘要　　運動資訊在生活上的應用非常廣泛，網路與數位科技的普及讓運動資訊在處理、傳遞與運用上變得更為方便，但是數位科技要有效地導入運動資訊應用的領域，需要跨領域的合作與知識的整合，目前還比較少有這樣的努力，所以國內運動資訊與相關應用的發展尚無法蓬勃。　　本研究提出一個一般化的運動資訊參考模型(generalized sports information reference model)，由此衍生出記錄運動資訊的系統架構，同時在這樣的基礎上與運動競賽的各種類型資料進行連結，產生各種潛在的應用，包括教學、訓練、運動科學及提供閱聽群眾的服務等。　　本研究也以從MVC衍生出的「空間-角色-事件」概念，來敘述所選擇建立知識本體與塑模的棒球運動，並透過統一塑模語言對既有資料進行資料採礦工程，以期發展出符合標準樣式的各種UML模型圖，作為設計開發軟體系統之參考架構。關鍵詞：運動資訊、數位學習、資料採礦、統一塑模語言、MVC

關鍵字

運動資訊；數位學習；資料採礦；統一塑模語言

並列摘要

ABSTRACT 　　Sports information has become easily accessible with the help of advancing digital and network technologies. The processing, delivery, display and use of sports information provide a solid base for all kinds of potential applications. However, the introduction of technologies into the area of sports information requires cross-discipline collaboration as well as integration of professional knowledge. 　　In this paper, a generalized sports information reference model is proposed to provide a high-level abstraction for different sports. We are then able to come up with a system architecture that records sports information and establishes links to sports video archives. The outcome contributes to applications in areas such as sports training, sports science, etc. 　　In this research, the generalized reference model for sports information is based on the space-actor-event triad; an innovative concept derived directly from the MVC (model-view-controller). The generic nature of the model makes it perfect as a basis for constructing a shared understanding of the sports domain; i.e., ontology for sports. 　　On the other hand, the model can also be specialized to describe a specific sport; e.g., basketball, badminton, etc. We chose baseball as a concrete instance for applying our model. Running through the analysis and design phase, the UML diagrams and relational tables emerge for our baseball information system. The catch lies in the metadata introduced in the process. The metadata is useful for accurate navigation within the space of sports information. The metadata along with the recorded information lay out the foundation for the sports knowledge base. Keyword：sports information、data mining、UML、MVC