I-Corps: Software that predicts which undergraduates are at risk of dropping out or requiring more than four years to graduate

点击翻译按钮获取中文摘要

The broader impact/commercial potential of this I-Corps project is the development of an artificial intelligence (AI)-based technology to enhance the understanding of college student success and how to help students align their financial resources and academic plans. The proposed technology advances techniques in data engineering, machine learning, and visualization, and may enable colleges and universities to leverage their massive student academic and financial records to support students’ success in more sophisticated, targeted capacities. The proposed technology may help higher education institutions achieve strategic goals for increasing graduation rates, improving time-to-degree completion, and eliminating equity gaps in graduation outcomes for low-income students. Students using the proposed technology may benefit from insights gained into the runway of eligibility for federal financial aid, allowing students to make more informed and efficient degree planning choices, graduate before exhausting lifetime aid eligibility, and limit their educational expenses and debt by reducing their overall time to degree. More individuals may earn baccalaureate degrees in less time and with less overall expense and student debt, and be ready to contribute better to society.This I-Corps project is based on the development of an intelligent system that incorporates financial, academic, and demographic data to accurately predict whether undergraduates are likely to graduate from 4-year institutions and if they will graduate within four years or more. The proposed technology has been designed to identify students who are at increased risk in terms of their financial health, including students on a trajectory to exhaust lifetime-limited federal financial aid resources before graduating. In addition, the developed scripts for data cleaning and feature engineering may be easily changed to fit a given university’s data models. Undergraduates may be able to use the proposed advanced web and mobile technologies to wisely plan their academic coursework in order to graduate before their financial aid runs out, resulting in improved academic outcomes. A full stack model of development characteristics includes software with layers developed based on the business requirements and processes while complying with flexible technology capable of working alone or integrating with existing enterprise software, secure integration of machine learning algorithms through microservices, and the ability to scale to large and small school sizes and integrate with different enterprise information systems.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.