Competition-based many-to-many academic collaborator recommendation

Masters by Research

Competition-based many-to-many academic collaborator recommendation 


We are witnessing scientific collaboration has become of great importance to the scientific community in undertaking research. However, it is difficult and time-consuming for scholars to find relevant collaborators because of the rapid rise of scholarly data and new scholars ceaselessly join academia. The recommendation generated by most existing methods is locally optimised. In contrast, this project will focus on globally-optimised collaboration based on completion. In reality, due to the limited time and collaboration ability of collaborators, scholars may need to compete with each other to win collaboration opportunities. This project will explore matching optimisation mechanisms such as market matching theory in order to iteratively recommend a collaborator with the highest match to each target scholar based on competition. The competition will be a two-way process, i.e., the set of targets will compete with each other to win their best match from the set of collaborators and vice versa. This will, in turn, maximise the overall benefits of the collaboration. The expected outcome of this project is a novel solution for group collaborator recommendation (many-to-many collaboration matching), which will result in globally-optimised collaboration recommendation.

Supervisory team

Principal Supervisor: A/Prof Feng Xia

Co-supervisor: Dr Nargiz Sultanova