A Hybrid Knowledge and Model Approach for Reviewer Assignment

Abstract

In R&D project selection, experts (or external reviewers) always play a very important role because their opinions will have great influence on the outcome of the project selection. It is also undoubted that experts with high expertise level will make useful and professional judgments on the projects to be selected. So, how to assign the most appropriate experts to the relevant proposals is a very significant issue. This paper presents a hybrid knowledge and model approach which integrates mathematical decision models with knowledge rules, for the assignment of experts to review of R&D project proposals. The approach can be applied to government funding agencies in China and other countries.

Reviewer Assignment Problem: A Systematic Review of the Literature

“…), others try to solve the assignment problem by optimizing objectives under some constraints (Cook et al, 2005;Li et al, 2008;Tang et al, 2010;Tang et al, 2012;Daş & Gökçen, 2014;Guo et al, 2018;Boehmer et al, 2021, etc.). Also, there are studies dealing with all stages of the RAP and designing it as a reviewer recommendation system or decision support system (e.g., Tian et al, 2005;Janak et al, 2006;Sun et al, 2007;Conry et al, 2009;Xu et al, 2010;Silva et al, 2014;Liu et al, 2016;Hoang et al, 2021).…”

On optimization of expertise matching with various constraints

“…The problem has attracted considerable interest from different domains. For example, several works have been made for conference paper-reviewer assignment by using methods such as mining the web [1], latent semantic indexing [2], probabilistic topic modeling [3] [4], integer linear programming [5], minimum cost flow [6] and hybrid approach of domain knowledge and matching model [7]. A few systems such as [8][9] [10] have also developed to help proposal-reviewer and paper-reviewer assignment.…”

Expert group formation using facility location analysis

“…This problem can be considered as an extension of the paper-review assignment problem. While initial approaches for paper-review assignment (Hettich & Pazzani, 2006;Sun, Ma, Fan, & Wang, 2007) concern only with the assignment of a paper to the relevant reviewers and ignore the coverage, confidence and load balancing conditions, Karimzadehgan and Zhai (2012) introduced the problem of multi-aspect constraint paper-review matching. They proposed an integer linear programming (ILP) method for this problem which is demonstrated in Algorithm 3.…”