New article on network sampling
The article has been published on Physica A by Xiao-Ke Xu, formally a postdoctoral fellow of the Web Mining Lab and now a Professor of Information and Communication Engineering at Dalian Minzu University, and Jonathan Zhu of the Lab. They proposed a new sampling method, called Self-Adjustable Random Walk (SARW), and empirically evaluated the quality of SARW against several prevailing sampling methods, including uniform, breadth-first search, random walk (RW), and revised RW sampling. The results show that SARW has been able to generate unbiased samples of OSNs with maximal precision and minimal cost.
Social media sampling has been an active line of research at the Lab. Previously, Jonathan and other members published an article on how to draw uniform samples from blogs, microblogs, etc., under the name of Random Digit Search (RDS) (Zhu et al., 2011).