Liu Yuanyuan, Kong Wenting, Merve Kaygac
School of Foreign Studies, China University of Petroleum (East China), Qingdao, Shandong, China.
School of International Education, China University of Petroleum (East China), Qingdao, Shandong, China.
Arq Bras Oftalmol. 2025 Jan 27;88(3). doi: 10.5935/0004-2749.2024-0269. eCollection 2025.
This article systematically reviewed 327 documents in the core collection of the Web of Science database regarding ChatGPT applications in the writing domain. This study aimed to comprehensively assess the latest progress and potential applications. ChatGPT demonstrates significant potential in overcoming writing anxiety, improving writing efficiency, generating initial scientific papers, and assisting researchers and students in giving feedback. However, it still faces significant challenges in data accuracy and the ethics of generated content, including inaccurate or outdated information, plagiarism risks, gender or race biases, etc. Authorship verification is particularly important for academic writing and publishing as it relates to objectivity, accuracy, and fairness. Future studies need to explore how to address these challenges through improvements at the technical and policy levels, ensuring that ChatGPT promotes the sustainable development and application of academic writing while adhering to ethical standards.
本文系统回顾了科学网数据库核心合集里327篇关于ChatGPT在写作领域应用的文献。本研究旨在全面评估其最新进展和潜在应用。ChatGPT在克服写作焦虑、提高写作效率、生成初步科学论文以及协助研究人员和学生提供反馈方面展现出巨大潜力。然而，它在数据准确性和生成内容的伦理方面仍面临重大挑战，包括信息不准确或过时、抄袭风险、性别或种族偏见等。作者身份验证对于学术写作和出版尤为重要，因为它关乎客观性、准确性和公正性。未来的研究需要探索如何通过技术和政策层面的改进来应对这些挑战，确保ChatGPT在坚持道德标准的同时促进学术写作的可持续发展和应用。
