面向文本的情感迁移分析策略

Text-Based Emotion Transformation Analysis

情感的迁移变化是外界刺激与个体感知共同作用的结果,本文着重就个体感知进行了探讨。首先就文本中情感的迁移规律进行了分析,讨论了心情与人格要素对情感迁移的影响。在此基础上,采用将语言特征与情感间迁移规律相结合的方法,通过机器学习实现了文本的情感分类。实验结果表明,情感分类的精确率相对传统方法提高了9.21%,方法的有效性得到了证明。

The emotion transformation is the result of the individual perception and stimulus, while the paper focuses on the individual perception. This paper firstly analyses the emotion transformation law and discusses the effect of mood state and personality factors. Then this paper presents an emotional automatic classification method which combines the language features and the emotion transformation law to identify the emotion agent utilizing the word features and the relationship among the contexts in emotion expressing sentences. The experimental results show that emotional automatic classification has obtained excellent effect which improves the accuracy by 9. 21%. The effectiveness of the method is proved in the paper.