Researchers at the College of Information and Computer Sciences (CICS) are part of a team creating a tool to analyze the profile and skills of workers, and then recommend training and education to help them grow in their career path. Funded by a grant from the National Science Foundation’s (NSF) Convergence Accelerator program, researchers from psychology, sociology, economics, and computer science will collaborate to develop algorithms and software to help companies and workers be successful in an ever-evolving workplace.
The Convergence Accelerator program supports multidisciplinary research teams to work on projects that will help companies apply “big data” in ways that enhance the lives of workers. Over 40 awards totaling $39 million were given to teams across the country to work on projects supporting two of NSF’s “10 Big Ideas”: Harnessing the Data Revolution and Future of Work at the Human-Technology Frontier. These areas focus on open knowledge networks to pool together many types of information, artificial intelligence to connect workers with jobs of the future, and innovative approaches to support workers seeking to upscale their skills.
“Technology and automation can lead to greater inequality among people and increased demands on workers. This NSF project will gather extensive data on workers, develop an infrastructure for data analysis, and support a deeper understanding of changes in the workforce based on advances in technology,” said Beverly Woolf, CICS research professor and principal investigator on this project.
The UMass team will collaborate with Stanley Black & Decker, the world’s largest tool manufacturer, to analyze anonymized data about workers’ skills, education, and career path, as well as information about job postings for open positions. In addition, the team will work with the city of Holyoke and MassHire Holyoke to analyze anonymized data about job seekers. The end goal is to create a tool where job seekers input their skills, interests, work experience, education, etc., and receive a skill assessment, recommendations about job opportunities, and suggestions for future training and education.
“Prior work mostly studied the problem of matching workers to jobs based on their current skill set, but has not studied the impact of upskilling and lifelong learning efforts,” said Andrew Lan, CICS assistant professor, and one of the key researchers on the project.
The researchers are committed to creating an equitable system, by focusing on fairness and equity at every stage of development. According to Shlomo Zilberstein, CICS professor and associate dean of research and engagement, “There are always unforeseen ways in which a system like this could end up being unfair. Fairness has many flavors -- there is no one box to check to make something fair. The way to avoid that is to give equitability a lot of thought at every step. ”
The team takes pride in how this work fits in with the CICS “computing for the common good” effort. “AI is not a technology that creates or solves problems by itself. It all depends on how you apply it -- if you apply it one way it can create problems, and if you apply it another way it can solve problems. We want to be part of the solution,” says Zilberstein.
The UMass collaborators on this research are Andrew Lan, Beverly Woolf, and Shlomo Zilberstein, CICS; Tom Juravich, sociology; Andrew Cohen, psychological and brain sciences; and Ina Ganguli, economics.