네트워크 분석과 토픽 모델링을 활용한 국내 임상간호사의 경험과 고충 분석: 온라인 커뮤니티 게시글을 중심으로

Abstract

Purpose: To describe clinical nurses’ experience and grievance in an online community using a co-occurrence network and topic modeling. Methods: We analyzed posts from Nurscape, which is the largest online community for nurses in Korea. After extracting posts using web scrapping, text preprocessing was done to detect nouns. In a visualization phase, co-occurrence network analysis and latent dirichlet allocation-based topic modeling were applied. Results: A total of 13,200 posts were analyzed. The co-occurrence network’s core keywords were newly graduate nurse, general ward, career, turnover, and grievance. The topic modeling showed four topics: (1) ‘Clinical practice-related difficulties’ described clinical hardships, such as the heavy workload of nurses; (2) ‘Concerns about resignation’ incorporated keywords asking for advice on resignation; (3) ‘Searching for information on employment/reemployment’ focused on the working conditions or working climate of a specific hospital; and (4) ‘Organizational action call’ captured the voices urging organized actions to improve nurses’ work environment. Conclusion: Clinical nurses share experiences through the online community and seek advice or information and urge organizational action. Professional nursing organizations should identify and deal with problems that nurses are currently facing. The results of this study can contribute to establishing the policy direction of nursing organizations. © 2021 Korean Academy of Nursing Administration