An integer linear programming model of reviewer assignment with research interest considerations

In the regular work process of peer review, editors have to read and understand the entire set of submissions to choose appropriate reviewers. However, due to a large number of submissions, to select reviewers manually becomes error-prone and time-consuming. In this research, a framework that considers different indispensable aspects such as topical relevance, topical authority and research interest is presented and, an integer linear programming problem is formulated with practical considerations to recommend reviewers for a group of submissions. Specifically, the topical relevance and the topical authority are utilized to recommend relevant and accredited candidates in submission-related topics, while the research interest is to exam the willingness of candidates to review a submission. To evaluate the effectiveness of the proposed approach, categories of comparative experiments were conducted on two large scholarly datasets. Experimental results demonstrate that, compared with benchmark approaches, the proposed approach is capable to capture the research interest of reviewer candidates without a significant loss in different evaluation metrics. Our work can be helpful for editors to invite matching experts in peer review and highlight the necessity to uncover valuable information from big scholarly data for expert selection.