Faculty of Social Sciences of UM recently held its fourth Computational Social Science Workshop. Prof. Haipeng Shen, Chair of Business Analytics and Innovation, University of Hong Kong gave the lecture to introduce the network centralitya popular indicator to assign numbers or rankings to nodes within a graph corresponding to their network position, which is often used in regression models to simulate the effect of networks on an outcome variable of interest.

Prof. Richard Weixing Hu welcomed Prof. Shen and pointed out that in today’s continuous practice and application of AI technology, the combination of big data and social science is of great significance.

Prof. Shen also introduced a new method for estimating centrality of supervised networks (SuperCENT), which produces better estimates of both centrality and the effect of network. He presented the advantages of SuperCENT in predicting currency risk premiums based on global trade networks.