텍스트마이닝을 활용한 전공경력 수준에 따른 챗GPT 활용 수업 경험 분석: 전문대학 1학년과 4학년(전공심화)을 중심으로 = A Text mining analysis of ChatGPT-based learning experiences by level of major-related experience: Focusing on first-year and fourth-year (advanced track) students in a junior college

다국어 초록 (Multilingual Abstract)

While learners’ learning experiences with ChatGPT in educational settings may vary depending on their level of major-related experience and type of institution (e.g., four-year university vs. junior college), few studies have examined these differences. This study aims to compare the ChatGPT-based learning experiences of first-year and fourth-year(advanced track) students at a junior college, categorized respectively as novice and advanced learners. A total of 266 reflective essays written by 56 first-year students and 38 fourth-year students who participated in a class during the second semester of 2024 were analyzed using text mining techniques, including Latent Dirichlet Allocation (LDA) and Social Network Analysis (SNA). The analysis revealed that, first, 3 topics were extracted from each group. First-year students’ experiences were categorized into: (1)using ChatGPT in the learning process, (2)supporting project tasks, and (3)perceived usefulness and inconvenience in Q&A interactions. Fourth-year students’ reflections revealed: (1)evaluating information reliability, (2)using ChatGPT as a collaborative problem-solving tool, and (3)generating ideas for practical application. Second, Keyword network analysis showed that for first-year students, “GPT” and “thought” were closely connected, with “use,” “utilization,” and “activity” also associated. For fourth-year students, “information,” “thought,” and “GPT” were tightly connected, with “activity” and “utilization” forming additional links. These findings indicate differences in experiences depending on level of major-related experience and suggest that instructors should design teaching methods that reflect learners’ varying levels when incorporating ChatGPT into class.