Abstract: Large networks are becoming pervasive in scientific applications. Statistical analysis of such large networks is prohibitive due to exorbitant runtime and high memory requirements. We propose a subsampling based divide-andconquer algorithm, SONNET, for community detection in large networks. The algorithm splits the original network into multiple subnetworks with a common overlap, and carries out detection algorithm for each subnetwork. The results from individual subnetworks are aggregated using a label matching method to get the final community labels. This method saves both memory and computation costs significantly as one needs to store and process only the smaller subnetworks. This method is also parallelizable which makes it even faster.

Key words and phrases: Community detection, computational efficiency, degree corrected blockmodel, spectral clustering, stochastic blockmodel, subsampling.