Title: A network compression approach for quantifying the importance of temporal contact chronology

Abstract: Studies of dynamics on temporal networks often look at the overall impact of temporal structural changes, falling back on static networks when the changes are slow, or on average annealed networks when the changes are fast. In between these limits, we propose a method to quantify the importance of chronology in temporal network data by using an epidemic spreading approach and looking at the commutative property of consecutive snapshots. We use this method to reduce the number of snapshots in real sequences of contact data by algorithmically compressing consecutive snapshots. We find that the framework succeeds in mimicking the fully temporal dynamics, and can also be used to assess the sensitivity of the data to temporal variations.