Scholars@Duke Collaboration Finder Using the Power of AI

Description

Scholars@Duke is a research discovery system featuring the research, scholarship, and activities of Duke faculty, graduate students, and academic staff. The system has a significant amount of information about scholarly activity at Duke and provides many direct relationships between faculty, such as those co-publishing a paper or on the same grant -- but what about the non-direct relationships that can be gleaned from the information in Scholars?

A team of students will explore various ways of using artificial intelligence (AI) to look for additional relationships between researchers at Duke. Using Scholars@Duke as the data source, the team will experiment with various pre-trained AI models, both commercial and open source, to find similarities between researchers that could help them identify new collaborators.

Watch the team's final presentation on YouTube.

Image

Team

Members

• Alexandra Doss
• Melina Marjani
• Rachel Kurapati
• Yulia Savine

Leaders

• Richard Outten
• Erik Daughtrey

Technical Resource

• Dauren Bizhanov

Project Facilitators

• Jaelyn Cuellar
• Mariam Gvenetadze