Presentation Information
[SS-2-01]Design and Prototype of a Bidirectional AI Matching System for International Students and Japanese Universities
*Shuichi Ito1, Xinli Li1, Toshiki Tagaya1, Yuxuan Cheng1, Kiyotaka Takahashi2 (1. Global Education Service CO.,LTD., 2. Kansai University of International Studies)
Keywords:
AI Matching,Student Recruitment,Bidirectional Matching,Database,Prototype Development
受講者に求められる 事前の知識・経験等
なし
受講者が受講前に取り組む 事前課題等
なし
概要
As competition for international student enrollment intensifies among Japanese universities, prospective students often face significant information asymmetry. This session opens with a report from Kansai University of International Studies on the current challenges of international student recruitment and management from the university's perspective. We then present findings from a survey of Chinese high school students enrolled in GES-operated preparatory schools and international education staff at Japanese universities, identifying three structural challenges: difficulty reaching prospective students, inefficient applicant management, and student-institution mismatches. We introduce a prototype bidirectional AI matching system developed to address these challenges. Students input their profile—including JLPT level, GPA, English proficiency, desired faculty, and budget—whereupon Claude AI scores their fit across ten axes and presents the top three universities with radar charts. Universities can adjust axis weightings to filter candidate students. We argue this bidirectional scoring approach represents a novel contribution to the field. Looking ahead, we are preparing to conduct a full PoC using real student data from two GES-operated Japan-study preparatory schools in Beijing and Shanghai, leveraging a network of 24 universities for which GES has been entrusted to operate as their China liaison office. Through co-presentation, we invite discussion on system design, current limitations, and future directions.