Toward a Better Measure of Business Proximity: Topic Modeling for Industry Intelligence

Table A. ERGM Notations Network Graph Y, Yij a random network graph matrix, its i,j element Y-ij all elements except i,j y the set of all possible graphs for a fixed set of notes y, yij y, yij a realization of the random network graph and its i,j element zk(y) a statistic of network graph y Network Statistics t total number of edges d2 number of nodes which have at least 2 edges hs number of edges within state s hc number of edges within category c pg sum of geographic proximity over all edges ps sum of social proximity over all edges pf sum of investor proximity over all edges pb sum of business proximity over all edges Nodal Characteristics: si state where i’s headquarters is located ci category to which i belongs Dyadic Characteristics pg,ij geographic proximity of i and j ps,ij social proximity of i and j pf,ij investor proximity of i and j pb,ij business proximity of i and j