Weight Prediction on Missing Links in Social Networks -- a Cross-Entropy-Based Approach --

A social network (SN) is a group of actors and their mutual relations. Sociologists try to answer the question why networked actors in our society aremore successful than others and how this networking works. Directed or undirected graphs, hyper- or multigraphs are a suitable means to visualize social relations. Social networks with directed and weighted links among actors need sophisticated instruments for analyses. We model these links as probabilistic conditioned propositions. Then for any actor i the model permits the estimation of transfer probabilities to all actors j, may they be linked to i or not. When future sociological research wants to interconnect missing links, some of the respective weights cannot be chosen at will but must fall in certain intervals. They must be in accordance with former conditional-logical net structure. To achieve this goal, cross-entropy-driven knowledge bases are applied. For a middle-size network we demonstrate the new findings.