Evaluating Text Summarization Tools for Educational Applications: A Study of Copilot, ChatGPT, and Quillbot
Thuong Hong Thi Nguyen, Vi Loi Truong, Ngan Tran Huynh Chau · 2025
The aim of this study is to analyze and evaluate the role of text summarization software (TSumSf) in modern education, aiming to improve learning efficiency and save time for learners. The study uses the BERT algorithm to compare the summarization capabilities of tools such as Copilot, QuillBot, and Chat GPT and compare the similarity (Sim) to the original text while providing information on condensation, accuracy, and time-saving ability. The results show that Copilot is suitable for abstract natural science and computer science texts that need to condense the content, while QuillBot is suitable for natural science articles. Chat GPT provides an overview, suitable for journalistic, literary, and technical-scientific texts. The choice of text summarization tools should be based on the specific purpose of use, thereby best supporting learners in accessing and processing information. Following this study, it was found that the summary ratio ranged from Sim = 0.83 to Sim = 0.97. In addition, the study also highlighted the importance of user perception of the usefulness and effectiveness of these tools, along with reliable infrastructure and seamless integration into the educational system.