This is an implementation of the temporal centrality statistics proposed by Hyounshick Kim and Ross Anderson on their paper "Temporal Node Centrality in Complex Networks" available at http://journals.aps.org/pre/abstract/10.1103/PhysRevE.85.026107.
Disclaimer:
I was not able to implement the temporal betweenness centrality exactly as stated on their paper, mainly because I did not understand the way the authors compute the normalization constant.
In this implementation, if there is a path that requires waiting at a particular node for k timesteps, the path is only counted once instead of k times as in the paper. The normalization constant for the temporal betweennes is (0.5*(n-1)*(n-2)*m) where m = timestemps in the temporal graph, n = number of vertices in the static graph.