Blockchain-based Reviewer Selection
Jianmin Wen, Sathiamoorthy Manoharan, Xinfeng Ye · 2022
Universities require their students to undertake projects as partial fulfilment of the students' degree. These projects typically go through an independent external review process so that academic standards are upheld. Finding appropriate external reviewers with the domain knowledge is a time-consuming process. In this paper, we propose a system that helps advisors find examiners for their students' projects. The system uses machine learning techniques to identify the subject areas of the students' projects from the text of students' dissertations. It then matches scholars who have research interests in the identified subject areas. The system is implemented using a blockchain and aims to provide a transparent environment that allows scholars to freely participate in the system. In this project, we collected a set of papers from several top journals and conferences. Using this dataset, we have evaluated the performance of several classification algorithms, e.g., multinomial Naive Bayes, K-Means Clustering with multinomial logistic regression and Labeled-LDA. The experimental study shows that the weighted multinomial Naive Bayes model performs best among the three models in terms of accuracy and computational time.