Weighting Assessment of the Effect of Chinese State-Changing Words on Emotions

Chia-Yueh Chang,Meng-Ning Tsai,Yao-Ting Sung,Shu-Ling Cho,Hsueh-Chih Chen
DOI: https://doi.org/10.1007/s10936-023-09986-9
Abstract:Past studies of sentiment analysis have mainly applied algorithms based on vocabulary categories and emotional characteristics to detect the emotionality of text. However, the collocation of state-changing words and emotional vocabulary affects emotions. For example, adverbs of degree strengthen emotions, and negative adverbs reverse emotions. This study investigated the weighted effect of state-changing words on emotion. The research material comprised 73 state-changing words that were collocated with four emotions: happiness, sadness, fear, and anger. A total of 84 participants participated in the vocabulary assessment. The results revealed that state-changing words could be classified into four types: intensifying, weakening, neutralizing, and reversing. In a comparison of the weighting factors among emotions, the weighting effect of the same state-changing word in the positive emotion category was particularly evident. The results could serve as a reference for follow-up studies on detecting emotions in text.
What problem does this paper attempt to address?