藝術的創作是一種從無到有的過程,裡面包含了創作者的觀念、想法及獨特 的美感,一般認為創作是沒有拘束的,但其實還是有隱約的規則與創作者的風格 存在。目前作曲工具大多以音訊處理或五線譜為設計介面,造成學習入門困難; 而一般大眾也很難理解隱藏在準則式作曲背後的作曲知識。 因此,本研究利用本體論將作曲的知識具體化,透過資料的分析整理,建立 出音樂作曲知識本體論的作曲本體論,提出利用此本體論將準則式作曲的物件貼 上標籤,訂立物件的名稱、內容屬性及與上層物件的關連,將隱藏的作曲知識具 體化。另外捨棄傳統的記譜方式,利用視覺與聽覺的共感覺,將色彩與聽覺的之 間關連音樂元素,並對應到音樂作曲知識本體論上的音樂屬性。藉由繪圖過程中 的色彩、形狀大小及位置的遠近變化,對應到準則式作曲中的物件進行作曲。 在未來發展方面,提出此模式的應用領域,包含:「藝術教學領域」,在繪圖 的過程中同時學習音樂的相關知識;「音樂能力評量」,透過音樂知識具體化,藉 此瞭解學生的音樂知識水平如何;「音樂治療」,透過藝術治療中常用的美術與音 樂兩項藝術,以繪圖的方式偵測使用者的情緒反應,再利用後端的準則式作曲產 生適當的音樂,作為一種即時藝術治療的方式。
With the development of compter technology, the music composition by computer has been widely accepted. Algorithmic composition is a composition method that automates some digital music patterns by the algorithm design. However, the design knowledge in algorithmic composition is usually implicitly embedded in the programming code. Therefore, the high level music creation knowledge via algorithmic composition is difficult for beginners to learn. To represent the high level concept behind the program, we propose a Music Composition Ontology (MCO) which can meaningfully represents the name of object (music knowledge), the metadata of object and the relation with multiple level objects. With the modeling of MCO, the traditional clef can be represented by the high level metadata. Furthermore, by the theory of Synesthesia, we map the color and the sense of hearing to the metadata of MCO object. Through the different color, the form size and the distance of position, we aim to develop a new way for algorithmic music learning. To prove the feasibility of this concept, we develop the Kandinsky system and several experiments are done. Finally, by the MCO we proposed, several research domains which can be applied in the future are also discussed. .