摘要
针对目前互联网上使用的音乐检索系统是以文本信息为主要搜索依据,难以满足智能化检索需求的问题,提出了一种基于内容的音乐检索方法。在保留传统检索方法的基础上,将情感和场景需求引入到检索中,对基于语义的音乐检索方法及模型进行了相关研究。首先根据音乐所蕴含的意境及适合播放的场景进行分析,构建音乐领域本体;然后对用户提出的查询请求进行分词分析和词性标注,并进行语义分析;最后通过推理机对本体进行推理,检索出满足用户需求的音乐集。
To resolve the problem that the Internet music retrieval system currently used takes text messages as its main search objects and is difficult to meet demands for intelligent search,a content-based music retrieval method is proposed.Based on traditional retrieval method,the new method innovatively introduces emotions and scenes into the search and is a semantic retrieval system.First,the artistic conceptions which music contains and the scenes suitable for music to play are analyzed to build music domain ontology.Then,word analysis,POS tagging and semantic analysis are carried on based on the queries which users submit.Finally,the music collections which satisfy user＇s needs are retrieved by inference on the music domain.
出处
《重庆理工大学学报（自然科学）》
CAS
2011年第1期40-43,共4页
Journal of Chongqing University of Technology：Natural Science
关键词
语义检索
音乐
本体
自然语言
semantic search
music
ontology
natural language
