Leveraging AI for a Common Course Numbering System in California Community Colleges

• In Event: AI and Transfer Student Success: Learning From the Field

Abstract

This study investigates the application of artificial intelligence (AI) to streamline the process of determining course equivalencies across 116 California Community Colleges. We propose an AI-driven Common Course Numbering System that aims to assign identical course numbers to comparable lower-division courses across all California community colleges, facilitating smoother transfers to four-year institutions.
Transfer pathways are crucial for promoting social mobility and access to higher education, particularly for underrepresented groups. Numerous studies highlight that robust course equivalency agreements between “sending” and “receiving” institutions are key to increasing transfer rates and reducing excess credit accumulation. However, the current course equivalency process is resource-intensive and time-consuming.
We employed large language models to create vector representations of general education courses based on their titles and catalog descriptions from California community colleges. We also integrated techniques from Natural Language Processing that utilize distinctive catalog description styles from various campuses to allow similar courses across campuses to be positioned near each other in the vector space. K-means clustering was then applied to these vector representations to group adjacent courses, potentially enabling a common course numbering system. This approach has successfully cleared a feasibility stage of application to five sample community colleges and is in the process of being scaled up to all 116 campuses.

Authors

• Yerin Kwak, University of California - Berkeley

• Siddharth Jayprakash Adelkar, University of California - Berkeley

• Zachary A. Pardos, University of California - Berkeley