基于特征旋律挖掘的二阶马尔可夫链在算法作曲中的研究与应用
Research and Application on the Second-Order Markov Chains Based on Feature Melody Mining in Algorithmic Composition
Abstract
算法作曲是如今国内外许多研究者和作曲家都十分关注和颇感兴趣的一个新兴跨学科领域。本文首先针对算法作曲领域内已有的各类研究做了资源整合、归纳与知识分享工作,即对与该领域相关的文献进行了较为详细的分类介绍与优劣述评,并在此基础上,结合自己的探究与思考而尝试设计和提出一种新的用于音乐片段分析与特征旋律挖掘的ISM算法,即音程序列挖掘算法。该算法是从用于关联规则挖掘的Apriori算法思想中获得启发,并结合相关乐理知识和作曲家作曲的心理过程来进行设计的,它以挖掘到的音乐片段中出现次数最多并相对较长的音程子序列作为该音乐片段的特征旋律。 算法作曲领域可以大致分为两个方向,一个方向是如何用算法分析已有的... Nowadays, algorithmic composition has become a new interdisciplinary field, many researchers and composers both at home and abroad are very concerned about it and find it interesting. In the thesis, all kinds of research resources of algorithmic composition have been integrated, concluded and shared with others, namely related literature in this area are collected and introduced in a more detailed...