音乐的情感标签预测对音乐的情感分析有着重要的意义。该文提出了一种基于情感向量空间模型的歌曲情感标签预测算法,首先,提取歌词中的情感特征词构建情感空间向量模型,然后利用SVM分类器对已知情感标签的音乐进行训练,通过分类技术找到与待预测歌曲情感主类一致的歌曲集合,最后,通过歌词的情感相似度计算找到最邻近的k首歌曲,将其标签推荐给待预测歌曲。实验发现本文提出的情感向量空间模型和“情感词—情感标签”共现的特征降维方法比传统的文本特征向量模型能够更好地提高歌曲情感分类准确率。同时,在分类基础上进行的情感标签预测方法可以有效地防止音乐“主类情感漂移”,比最近邻居方法达到更好的标签预测准确率。
Abstract
Music emotion tag prediction algorithm plays an important role in music sentiment analysis. This paper presents a sentiment vector space model (s-VSM) based music emotion tag prediction algorithm. Firstly, we extract the emotion words to build the sentiment vector space model. Then, we use SVM classifier to generate training samples, and to get the collection which shares the same main emotion category with the predicted music. Finally, by finding the nearest k songs, we can get the emotion tag for recommendation. Experimental results show that s-VSM and the “emotional words-emotional label” co-occurrence based feature reduction method perform better than traditionally word-based vector space model in mood classification. Meanwhile, the emotion tag prediction based on the result of classification can effectively prevent the music “main mood drift”, thus achieving better tag predict accuracy than k-nearest neighbors method.
Key wordstag prediction; feature reduction; mood classification; sentiment vector space model
关键词
标签预测 /
特征降维 /
情感分类 /
情感向量空间模型
{{custom_keyword}} /
Key words
tag prediction /
feature reduction /
mood classification /
sentiment vector space model
{{custom_keyword}} /
{{custom_sec.title}}
{{custom_sec.title}}
{{custom_sec.content}}
参考文献
[1] X. Hu, J. S. Downie, C. Laurier, et al. The 2007 MIREX Audio Music Classification Task: Lessons Learned[C]//Proceedings of the International Conference on Music Information Retrieval, Vienna, Austria,2008:462-467.
[2] 夏云庆,杨莹,张鹏洲,等.基于情感向量空间模型的歌词情感分析[J].中文信息学报,2010,1(24):99-103.
[3] Hevner K. Expression in music: a discussion of experimental studies and theories[J]. J. Am. J. Psychiatry.1936:246-268.
[4] Dan Liu, Lie Lu. Automatic mood detection from acoustic music data[C]//Proceedings of the International Symposium on Music Information Retrieval, Baltimore, MD, USA, 2003: 81-87.
[5] Ogihara M. Content based music similarity search and emotion detection[C]//Proceedings of 2004 IEEE International Conference on Acoustics, Speech, and Signal Processing, Montreal, Quebec, Canada, 2004:17-21.
[6] Dan Yang, Won-Sook Lee. Music Emotion Identification from Lyrics[C]//Proceedings of 11th IEEE International Symposium on Multimedia, San Diego, Canada, 2009: 624-629.
[7] 陈若涵.以音乐内容为基础的情绪分析及辨识[C]//第二届电脑音乐与音讯技术研讨会, 台北, 2006:68-75.
[8] D. Yang, W. Lee. Disambiguating music Emotion Using Software Agents[C]//Proceedings of the 5th International Conference on Music Information Retrieval, Barcelona ,Spain, 2004: 52-58.
[9] C. Laurier, J. Grivolla , P. Herrera. Multimodal Music Mood Classification Using Audio and Lyrics[C]//Proceedings of the International Conference on Machine Learning and Applications, San Diego, Canada,2008:688-693.
[10] Y. H. Yang, Y. C. Lin, H. T. Cheng, et al. Toward multi-modal music emotion classification [C]//Proceedings of Pacific Rim Conference on Multimedia, Tainan,China,2008:70-79.
[11] 夏云庆. 歌曲情绪压力分析方法及系统: 中国,200910087827.X[P],2009,11,25.
[12] Thayer R. E. The biopsychology of mood and arousal[M]. Oxford University Press, 1989.
[13] 徐琳宏,林鸿飞,潘宇,等. 情感词汇本体的构造[J]. 情报学报, 2008, 27(2): 180-185.
[14] 孙守迁,王鑫,刘涛,等. 音乐情感的语言值计算模型研究[J]. 北京邮电大学学报. 2006, 11(29): 35-40.
{{custom_fnGroup.title_cn}}
脚注
{{custom_fn.content}}
基金
国家自然科学基金资助项目(60673039,60973068);国家社科基金资助项目(08BTQ025);国家863高科技计划资助项目(2006AA01Z151);教育部留学回国人员科研启动基金和高等学校博士学科点专项科研基金资助项目(20090041110002)
{{custom_fund}}
PDF(2750 KB)
