Expert recommendations based on link prediction during the COVID-19 outbreak

Abstract

Since the emergence of COVID-19, the number of infections has significantly increased. As of April 7, 8:00 am, the total number of global infections has already reached 1,338,415, with the number of deaths being 74,556. Medical experts from various countries have conducted relevant researches in their own fields and countries, and the development of an effective vaccine has been expected soon. Although some progress has been made in the development of therapeutic drugs and vaccines, interdisciplinary and cooperative studies are scarce. However, it is easy to form information islands and conduct repeated scientific research. To date, no therapeutic drug or vaccine for COVID-19 has been officially approved yet for marketing. In this article, the features of experts in cooperation networks, such as graph structure, context attribute, sequential co-occurrence probability, weight features and auxiliary features, are comprehensively analyzed. Based on this, a novel graph neural network + long short-term memory + generative adversarial network (GNN + LSTM + GAN) expert recommendation model based on link prediction is constructed to encourage cooperation among relevant experts in research social networks. Finding experts in related fields, establishing cooperative relations with them and achieving multinational and cross-field expert cooperation are significant to promote the development of therapeutic drugs and vaccines.