摘要
"萌"作为一种文化现象,在传播中常常诱发丰富的情绪体验。本研究立足于积极情绪理论与脑电情感计算技术,建立起积极情绪的十分类脑电预测模型,识别被试在观看"萌"视频时的情绪状态。结果表明"萌"视频诱发出多种积极情绪共现的情绪档案,其中最为突出的情绪包括爱、高兴、娱悦与宁静。此外,本研究展示的秒级时间精度的情绪体验识别系统有望在未来的传播实践中为媒介内容的"萌点"检测提供一种新的技术实现可能性。
“Moe”things often evoke rich emotional experiences.Based on the theory of positive emotions and affective computing techniques,the present study developed an EEG-based classifier for ten discrete positive emotions to identify the emotional states induced by“Moe”videos.The results showed that“Moe”videos induced multiple co-occurring positive emotions,such as love,joy,amusement,and serenity.In addition,the present study demonstrated an emotion recognition system with a second-level temporal resolution,which is expected to provide a promising technique for detecting the“Moe”points of media contents in future applications.
作者
胡鑫
蒋俏蕾
宋磊
张丹
Xin Hu;Qiaolei Jiang;Lei Song;Dan Zhang(Department of Psychology,School of Social Sciences,Tsinghua University;School of Journalism and Communication,Tsinghua University;China Animation Comic Game Group)
出处
《全球传媒学刊》
CSSCI
2021年第6期27-44,共18页
Global Journal of Media Studies
基金
国家自然科学基金(项目批准号:61977041)
教育部人文社会科学研究一般项目(项目批准号:17YJA190017)
多模态认知计算安徽省重点实验室(安徽大学)开放基金(项目批准号:MMC202001)资助。
关键词
萌
媒介效果
情绪档案
脑电
情感计算
Moe
Media Effect
Emotion Profile
EEG
Affective Computing
