Process Analytics Approach for R&D Project Selection

R&D project selection plays an important role in government funding agencies, as allocation of billions of dollars among the proposals deemed highly influential and contributive solely depend on it. Efficacious assignment of reviewers is one of the most critical processes that controls the quality of the entire project selection and also has a serious implication on business profit. Current methods that focus on workflow automation are more efficient than manual assignment; however, they are not effective, as they fail to consider the real insight of core tasks. Other decision models that analyze core tasks are effective but inefficient when handling large amounts of submissions, and they suffer from irrelevant assignment. Furthermore, they largely ignore real deep insight of back-end data such as quality of the reviewers (e.g., quality and citation impact of their produced research) and the effect of social relationships in project selection processes that are essential for identifying reviewers for interdisciplinary proposal evaluation. In light of these deficiencies, this research proposes a novel hybrid process analytics approach to decompose the complex reviewer assignment process into manageable subprocesses and applies data-driven decision models cum process analytics systematically from a triangular perspective via the research analytics framework to achieve high operational efficiencies and high-quality assignment. It also analyzes big data from scientific databases and generates visualized decision-ready information to support effective decision making. The proposed approach has been implemented to aid the project selection process of the largest funding agency in China and has been tested. The test results show that the proposed approach has the potential to add great benefits, including cost saving, improved effectiveness, and increased business value.