Reproducing feedback for environmental education in-class activities using large language models in virtual learning systems

A large language model (LLM) is an artificial intelligence program that understands, reads, interprets, creates, and generates new texts from dense datasets. The program can self-monitor with countless data points and easily perform text correction and transformation operations. Due to its self-monitoring capability, LLM can also easily track the potential outcomes of a sequence. This study aims to utilize this deep learning method, capable of understanding human language, in the field of education to facilitate the work of academics and students. The study was conducted with undergraduate students in the Faculty of Education as part of the Environmental Education course, thus testing the LLM model in the field of environment and sustainability, a topic of great need in recent times. Five environmental and sustainability themes were distributed to 39 students, and the texts were reorganized using the necessary commands with the LLM. The results were validated and verified by the researchers. This engaging study was conducted with enjoyment by the students. The students' re-evaluation of the reorganized texts proved that learning was achieved as intended. The model's ability to provide feedback and corrections to each text individually further contributed to facilitating the work of academics and providing diverse perspectives.