정서표현 글의 자동 분석을 위한 부정감성 사전 구축 방안 = Developing a Negative Emotion Lexicon for Automated Analysis of Expressive Writing

다국어 초록 (Multilingual Abstract)

This study aims to analyze the negative emotional vocabulary appearing in university students’ expressive writing and to construct a structured negative emotion lexicon that can be applied for educational purposes. Expressive writing reflects the writer’s inner emotions and experiences, and the negative emotions expressed therein serve as important indicators of the writer’s psychological and emotional state. Based on a dataset of 169 students’ autobiographical writing collected from a university-level academic writing course, a total of 639 negative emotional words were extracted, and 517 were finalized through expert review. Each word was then evaluated in terms of the intensity and proportion of six emotional categories—anger, anxiety, depression, sadness, loneliness, and frustration—using a modified semantic differential method and a 10-point rating scale. The results revealed distinctive patterns across emotional categories, with ‘anxiety’ and ‘depression’ showing the highest prevalence, while ‘loneliness’ appeared less frequently. This study demonstrates the possibility of quantitatively diagnosing learners’ emotional expressions and suggests the educational utility of a structured emotion lexicon in writing instruction. It also provides a practical foundation for integrating emotion analysis technologies with personalized feedback in emotion-based writing education.