arxiv.org
An arXiv paper presenting a large-scale analysis of the Tor network using a crawler called Mimir, which collects and visits linked content. The authors report a dataset of pages from more than 25k sites and examine network topology, replicated content and content types.