Finding Spread Blockers in Dynamic Networks

Social interactions are conduits for various processes spreading through a population, horn tumors and opinions to behaviors and diseases In the context of the spread of a disease or undesirable behavior, it is important to Identify blockers individuals that are most effective in stopping or slowing down the spread of a process through the population This problem has so far resisted systematic algorithmic solutions In an effort to formulate practical solutions; in this paper we ask. Are there structural network measures that are indicative of the best blockers in dynamic social networks? Our contribution is two-fold First, we extend standard structural network measures to dynamic networks Second, we compare the blocking ability of individuals in the order of ranking by the new dynamic measures We found that overall, simple ranking according to a node's static degree, or the dynamic version of a node's degree, performed consistently well Surprisingly the dynamic clustering coefficient seems to be a good indicator, while its static version performs worse than the random ranking This provides simple practical and locally computable algorithms for identifying key blockers in a network