Using Machine Learning and NLP to Understand Skills Development in the Undergraduate Curriculum – UROP Spring Symposium 2024

Using Machine Learning and NLP to Understand Skills Development in the Undergraduate Curriculum

Anirudh Brahmarouthu

Pronouns: he/him

Research Mentor(s): Annaliese Paulson
Research Mentor School/College/Department: Higher Education / Education
Program:
Authors:
Session: Session 1: 9:00 am – 9:50 am
Poster: 31

Abstract

My project is about using machine learning to interpret large sets of undergraduate curriculum and syllabi to analyze the various academic principles they prioritize. Through this, we are researching the link between the economic success that students achieve and the skills developed through their coursework. There has always been a debate on the usefulness of a well-rounded liberal arts education for all college students, regardless of major. For example, at U-M, we have the LSA distribution requirements so all students have to take a certain number of humanities and STEM courses regardless of their major. However, does this multidisciplinary education really advance the education of students? In the beginning of the project, we annotated large numbers of undergraduate curriculums and syllabi to differentiate different types of assignments, such as essays, exams, science labs, presentations, etc. Using these annotations, we are training machine models to replicate this process and identify what type of assignment syllabi at scale contain. By identifying the breakdown of assignment types certain students and majors have, we can then see the economic and labor success of these students after they complete college. Then we will be able to determine if having an interdisciplinary education actually benefits everyone, or if having a certain type of education that emphasizes only one or two concepts is more beneficial and leads to more success. We are currently still in the process of annotating and training our machine model with this information so we do not have any results. Our research will be able to quantify the “value” that different types of educations have in society, With our research, we will be able to help higher education institutions understand the true value of education and how to change their curriculums to optimize economic and labor success for their students.

Engineering, Interdisciplinary, Physical Sciences, Social Sciences

lsa logoum logo