Building a Funding Recommendation Engine from Scholars@Duke Data

Damaris Murry

Keywords: API, recommendation engine, MeSH, Library of Congress

TL;DR: The VIVO widgets API enabled a recommendation engine to be built using Scholars@Duke data

Abstract: Scholars@Duke is Duke University’s implementation of VIVO. The data provided in Scholars@Duke is shared widely throughout the university for websites, application development, reporting, visualizations, and more. One key feature in Scholars@Duke is the VIVO widgets API, which makes VIVO data available in an easy-to-consume JSON format. The widgets API helps to disseminate useful faculty and research information throughout our institution. Widespread usage of the widgets also reemphasizes to researchers the importance of maintaining their Scholars@Duke profile. The VIVO widgets are being used to match funding opportunities to Duke researchers. The recommendation engine, developed at Duke, makes use of subject headings from a researcher’s Scholars@Duke profile, which results in them receiving a personalized list of potential funding opportunities. In this talk, we will look at the VIVO widgets and its supporting documentation (API documentation, Terms of Use, and support policies) as well as the funding recommendation tool.