Improving the performance and scalability of graph anomaly detection with a new approach to edge significance for dynamic bipartite projections
Identifying edges significant to graph anomaly calculations by building on the binomial model for bipartite networks