Large Language Model in Higher Education: Leveraging Llama2 for Effective Academic Advising

Recent artificial intelligence (AI) developments have generated a new wave of technologies, with significant attention focused on large language models (LLM) providing users with information based on their input queries. This paper explores the application of LLM in the context of higher education, with the ultimate aim of supporting university students in consulting services. Despite numerous challenges due to the infancy of this technology, we believe that LLM-powered solutions in admission consulting can offer significant benefits to students. Compared to traditional consulting services, this new approach can enhance students' data confidentiality and provide them with a higher level of comfort when seeking information related to their academic performance and scores. Motivated by these potential benefits, this paper utilizes Llama 2, a well-known LLM with high linguistic analysis capability and accuracy. Llama 2 was fine-tuned using a dataset of admission-related queries from students at the International University - Vietnam National University, Ho Chi Minh City. Throughout the experiments and insights discussed, we aim to introduce our case study on how an LLM can be reconfigured and adapted for admission consulting in a higher education environment.