摘要
基于新闻有稿播音,采用多视角分析方式,探究言语情感生成及其衍化的可能机制。我们选取了与情感联系最为直接的三方面——认知评价、心理感受和生理状态,来构建言语情感描述体系,以期能够更为精准地刻画情感类型,反映情感的生成过程。在实验环节,组建了全部由播音专业高年级学生参与的测试队伍,通过对其心理感知实验数据的聚类、相关计算分析,形成了情感描述体系中的分级结构。
The generation and evolution mechanisms of speech emotion are explored based on a news broadcast corpus using a multi-perspective analysis approach.The most related three perspectives:cognitive evaluation,psychological feeling,and physiological state,are adopted to compose a transcription system,with which the refinement of speech emotion can be described more accurately while the process of speech emotion generation can be revealed clearly.Based on psychological tests,whose subjects are all senior broadcasting students,a hierarchical structure of the system is obtained via clustering and correlation analysis.
作者
高莹莹
朱维彬
GAO Yingying;ZHU Weibin
出处
《中国语音学报》
2013年第1期71-81,共11页
Chinese Journal of Phonetics
关键词
言语情感
情感产生
情感描述
Speech emotion
Emotion generation
Emotion transcription