An environment to foster scientific collaborations

Abstract

Collaboration is an integral part of modern scientific research. It is hard to imagine any modern scientific endeavor that does not involve a group of people. However, collaboration is frequently constrained to close partners or proximal groups. Finding new potential partners with whom to exchange ideas is an important step in establishing new working groups. We propose to use automated system to recommend potential collaborators, and present a system for this purpose. One of the problems involved is establishing criteria to select these potential partners. We address this issue by (1) automatically constructing users profiles from publicly available data to establish areas of expertise and (2) employing an analysis of users' networks to select partners with the potential for fruitful relationships.