http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
송영숙(宋永淑) ( Song Young-sook ),조원익(趙元翼) ( Cho Won-ik ) 한국어문교육연구회 2024 어문연구(語文硏究) Vol.52 No.1
본 연구에서는 사람과 인공지능 모델의 협업 시스템 구축을 통해 자동 화행 주석의 품질을 높이고자 하였다. 이를 위해 국립국어원 메신저 말뭉치(2022)에서 총 33,138 문장을 선별해서 사람과 GPT-4 모델을 통해 주석을 한 후 RoBERTa 모델을 사용해서 분류 정확도를 평가하였다. 사람이 주석한 메신저 말뭉치의 정확도 평가 점수는 85.50인데 비해 사람과 GPT-4 모델이 모두 주석한 경우만으로 정확도를 평가한 경우 GPT-4 모델이 가장 자주 틀리는 주석의 범주는 '냉소/유머'였다. 이렇게 불일치를 보이는 경우를 제외하고 모델과 인간의 주석이 일치하는 문장으로만 분류 모델을 통해 성능 실험을 했을 때 정확도 점수는 95.32로 매우 높은 정확도를 보였다. In this study, we aim at improving the quality of speech act annotation by identifying and analyzing the annotation agreement between humans and the GPT-4 model to identify the categories of speech acts that the GPT-4 model frequently misclassifies. For this purpose, we selected a total of 33,138 sentences from the National Institute of Korean Language’s Messenger corpus (2022). The annotation was performed by human annotators and the GPT-4 model, and the classification accuracy was evaluated using the RoBERTa model. The accuracy score for the human-annotated messenger corpus is 85.50, while the accuracy score for the cases where both humans and the GPT-4 model annotated identically is 96.32. In particular, the category of speech acts that the GPT-4 model most frequently misclassifies is “sarcasm/humor.” This is likely due to highly frequent expressions such as “ㅋㅋㅋ” (lol) in the messenger corpus, which causes the model to incorrectly analyze even “Statement” as “sarcasm/humor.” For example, even when a human annotator classified the utterance “I have snacks and ramen. ㅋㅋㅋ” as “Statement”, the GPT-4 model misclassified it as “sarcasm/humor.” In an experiment conducted using only the sentences where the model and human annotators agreed, the accuracy score reached 95.32. Instead of merely discussing annotator agreement for newly proposed corpora, it is necessary to distinguish speech acts that are difficult to annotate from those that are relatively easy. Furthermore, within these categories, it is essential to select sentences that are challenging to automati cally annotate and conduct research focused on more challenging areas.
간호대학생의 임상실습스트레스와 소진과의 관계에서 공감의 매개효과
송영숙(Song, Yeoungsuk) 한국간호교육학회 2018 한국간호교육학회지 Vol.24 No.4
Purpose: The purpose of this paper was to examine the relationships between clinical practice stress and burnout, and the mediating role of empathy in nursing students. Methods: A cross-sectional research design was employed. One hundred seventeen nursing students completed questionnaires on clinical practice stress, empathy (perspective-taking, fantasy, empathic concern and personal distress) and burnout in December 2017. IBM SPSS Statistics 23 was used and descriptive statistics, frequency, Pearson correlation coefficient and multiple regression were conducted. Baron & Kenny method and Sobel test were adopted for analysis of the mediation effect (personal distress of empathy). Results: The mean scores of clinical practice stress and burnout were 3.45 and 43.09, and perspective-taking, fantasy, empathic concern and personal distress of empathy were 2.67, 2.42, 2.64 and 2.19, respectively. The highest relationship between clinical practice stress and empathy was personal distress of empathy (r=.32, p<.001). Burnout was positively associated with clinical practice stress and personal distress of empathy (r=.22, p=.015; r=.51, p<.001). Personal distress of empathy demonstrated a complete mediating effect on the relationship between clinical practice stress and burnout (Z=3.22, p=.001). Conclusion: These results showed that decreasing personal distress of empathy is important for nursing students, and may help in reducing clinical practice stress and burnout.