The network community is a platform for people to communicate. In order to accurately analyze the emotions displayed in music community, this paper proposes a convolutional neural network classification model based on...The network community is a platform for people to communicate. In order to accurately analyze the emotions displayed in music community, this paper proposes a convolutional neural network classification model based on multi-dimensional emotions. Firstly, to solve the problem of feature extraction of emotion words under similar sentence patterns, it proposed a multi-emotion classification method and emotion vector splicing method that conform to music community emotion characteristics. Secondly, aiming at the coexistence of multiple categories of emotions in music comment text, it applied an emotional value measurement method based on music characteristics. Finally, the classification model was constructed with combining methods of emotion vector splicing and emotion value measurement. Through experimental analysis, this model is proved to have good performance in accuracy.展开更多
基金the National Natural Science Foundation of China under Grant No. 61672179, 61370083 and 61402126The Youth Foundation of Heilongjiang Province of China under Grant No. QC2016083+1 种基金the Fundamental Research Funds for the Central Universities under Grant No. HEUCF180606the Innovative Talents Research Special Funds of Harbin Science and Technology Bureau under Grant No. 2016RQQXJ128.
文摘The network community is a platform for people to communicate. In order to accurately analyze the emotions displayed in music community, this paper proposes a convolutional neural network classification model based on multi-dimensional emotions. Firstly, to solve the problem of feature extraction of emotion words under similar sentence patterns, it proposed a multi-emotion classification method and emotion vector splicing method that conform to music community emotion characteristics. Secondly, aiming at the coexistence of multiple categories of emotions in music comment text, it applied an emotional value measurement method based on music characteristics. Finally, the classification model was constructed with combining methods of emotion vector splicing and emotion value measurement. Through experimental analysis, this model is proved to have good performance in accuracy.
