Course selection plays a critical role in students’ academic success and satisfaction at university, necessitating effective support systems to assist them in making informed enrollment decisions. This study explores the development of a course recommendation system tailored to individual learning needs. The proposed system integrates two main approaches: one based on analyzing historical course registration records and academic performance of past students, and another leveraging content analysis of course syllabi. This concept is analogous to recommendation systems used in e-commerce, where suggesting relevant products based on customer behavior can significantly enhance user satisfaction. Applying a similar approach to course selection is expected to foster more personalized and effective learning experiences for students.

Chatbots, defined as interactive systems capable of understanding and automatically responding to user queries, are increasingly employed across various domains such as marketing, education, public services, and entertainment. Among popular chatbot deployment platforms, LINE (https://www.line.me/en/) stands out, especially in Japan, where it is widely used—90% of teenagers and 95% of individuals in their twenties actively use LINE, with monthly active users reaching 96 million as of December 2024. Organizations across sectors are leveraging LINE's chatbot features to enhance user engagement and communication.

This paper proposes the integration of a LINE-based chatbot interface into the existing web-based university course enrollment recommendation system. The core of the system comprises a set of programs that retrieve suitable course suggestions from a database. By transitioning from a traditional web interface to a LINE chatbot, the system aims to improve accessibility, usability, and student satisfaction. The interaction design has been fundamentally restructured to accommodate conversational user flows. While the web version allows students to input keywords and receive course recommendations in a single step, the chatbot version enables iterative dialogue, allowing users to refine their preferences and obtain more tailored course suggestions through continuous interaction.

Keywords: Course enrollment recommendation system, affinity analysis, apriori algorithm, chatbot.