音樂創作的未來發展,與文化創意產業休戚相關。為了使音樂創作能給更多的使用者創作使用,本研究分析自動作曲之特性,提出自動作曲在旋律產生方面需要的相關音樂的參數,希望可以對自動作曲的音樂分析上能更有效率。因此本研究把旋律曲線(Melodic Contour)中的Parsons Code對旋律做分析並加入大跳 (jump) 與級進(stepwise) 的音程藉此找出其中的音樂參數,在借由這些參數應用於音樂搜尋與自動作曲上。本研究的實驗結果,在提高解析度的音樂搜尋上,都有比Parsons code 更快的搜尋到想要找的音樂,在使用性的評估中,有效性上使用者還是覺得Parsons code比較簡單學習但在迅速性、功能性上提高解析度的評價都優於Parsons code。在本文所提出的自動作曲系統方面整體評價平均值72.8,表示大多數的使用者對於自動作曲的評價是不錯的,其中又以不懂音樂的使用者評價較優。
The future development of the music composition is used for culture and creative industry. In order to let more users compose their own music easily, this research proposes an innovated way to analyze the property of Automated Music Composition, determining the necessary features for algorithm composition, and makes Automated Music Composition of music analysis be more efficient. This research involves more detailed intervals in both jump and stepwise. In this thesis, we use Parsons Code for Melodic Contours to analyze the melody and include jump and stepwise to find out the musical features. Finally algorithm composition is applied to music composition. The experimental result shows the Melodic contour in a higher resolution is faster than the original Parsons code. We apply various methods of system usability scale, usability testing, effectiveness of using Parsons code is relatively simple, but functionality and efficiency of the Melodic contour in a higher resolution is superior than Parsons code. In our proposed Automated Music Composition, the overall average evaluation point of 72.8, shows that most people can accept the Automated Music Composition system.