Similarity Based Compression Ratio for Dynamic Network Modelling

Dynamic network modelling of timely evolving complex systems allows to discover emerging properties of realworld facts. The main issue of such modelling is determining the proper time intervals, a.k.a. window size, for each member of network. In this work, we propose a new network similarity based compression ratio for measuring the properness of studied window size. Besides, we show that a more informative dynamic network with a less noisier structure can be extracted by using a window aggregation strategy. The results on Enron, Haggle Infocom and Reality Mining data sets reveal that the proposed compression ratio is more effective for finding best window size than baseline and aggregation strategy allows to capture important time-dependent events which might be hidden in noise when using constant windows.