摘要
传统的基于文本信息描述的音乐检索技术已经无法满足人们对检索智能化的需求,于是产生了基于内容的音乐检索方法。在此基础上将情感需求引入到检索中,对基于情感的音乐检索方法及模型进行了相关研究。首先构建了音乐情感空间来获得用户的情感描述;然后通过对情感音乐模型进行定义提出了情感音乐模板库,以得到满足用户情感需求的匹配模板;最后,在此基础上提出了基于情感音乐模板的音乐检索系统模型,力求探讨出一种基于情感的有效检索方法。
The traditional information retrieval technology which is based on text has been unable to satisfy the need for more and more music information. So content-based music retrieval technology was developed, which applies audio characteristics. This paper researched the music retrieval method and model based on affection by importing affective requirements from users. We constructed a music affective space in order to acquire users' affective description; introduced an affective music pattern database through defining PDA model, so as to get a music pattern satisfied by users' affective requirements; we put forward a music retrieval system model based on affective music pattern, aiming to think out an efficient affective-based retrieval method.
出处
《计算机科学》
CSCD
北大核心
2009年第1期239-241,278,共4页
Computer Science
基金
天津市自然科学基金项目(编号:06YFJMJC00100)
天津师范大学青年基金(编号:52LJ89)资助