Dynamic network analytics for recommending scientific collaborators

Collaboration is one of the most important contributors to scientific advancement and a crucial aspect of an academic’s career. However, the explosion in academic publications has, for some time, been making it more challenging to find suitable research partners. Recommendation approaches to help academics find potential collaborators are not new. However, the existing methods operate on static data, which can render many suggestions less useful or out of date. The approach presented in this paper simulates a dynamic network from static data to gain further insights into the changing research interests, activities and co-authorships of scholars in a field–all insights that can improve the quality of the recommendations produced. Following a detailed explanation of the entire framework, from data collection through to recommendation modelling, we provide a case study on the field of information science to demonstrate the reliability of the proposed method, and the results provide empirical insights to support decision-making in related stakeholders—e.g., scientific funding agencies, research institutions and individual researchers in the field.


Author(s): Lu Huang, Xiang Chen, Yi Zhang, Yihe Zhu, Suyi Li, Xingxing Ni
Organization(s): Beijing Institute of Technology, University of Technology Sydney
Source: Scientometrics
Year: 2021

Leave a Reply

Your email address will not be published.