A Novel Local Propagation Based Expert Finding Method

Abstract

The number of academicians and academic studies in the world is increasing day by day. At the same time, with the development of science, the area of expertise has become even more important. Finding the right reviewer or jury in the analysis of the academic study is very important for the evaluation of the study. In the academic world, many studies have been done in the literature about finding experts. All of these studies deal with the studies of the candidates and the background information and metadata of the studies. Just looking at the background information and metadata of studies is not enough alone to find an expert in the academic world. Finding another way to understand expertise level of academics has become an interesting problem. Therefore, looking at co-authors to understand expertise level of reviewers is an excellent way. In this study, it is focused on both statistical approaches and propagation-based Improved Topical Affinity Propagation (I-TAP) approach to evaluate experts. In our method, expert candidates have been examined according to not only background information but also expertise of co-author relationship. The study was also scored with II-TAP and statistical methods to find reviewers. The proposed method is applied to a data set of 250,000 articles and the test results are presented comparatively in terms of F-Measure with other similar approaches.