Amazon product co-purchasing network and ground-truth communities

The Amazon product co-purchasing network and ground-truth communities dataset, part of the Stanford Network Analysis Project (SNAP), captures a network collected by crawling the Amazon website, based on the "Customers Who Bought This Item Also Bought" feature: an undirected edge connects products frequently co-purchased together, and each Amazon product category defines a ground-truth community. It includes the largest connected component of the co-purchasing network, ground-truth community assignments (connected components per category with at least 3 nodes), and the top 5,000 highest-quality communities, with 334,863 nodes, 925,872 edges, an average clustering coefficient of 0.3967, and a diameter of 44.

Use the following BibTeX citation for the SNAP network datasets:

@misc{snapnets,
  author       = {Jure Leskovec and Andrej Krevl},
  title        = {{SNAP Datasets}: {Stanford} Large Network Dataset Collection},
  howpublished = {\url{http://snap.stanford.edu/data}},
  month        = jun,
  year         = 2014
}