Walk, not wait: Faster sampling over online social networks

Azade Nazi, Zhuojie Zhou, Saravanan Thirumuruganathan, Nan Zhang, Gautam Das

Research output: Contribution to journalConference article

15 Scopus citations

Abstract

In this paper, we introduce a novel, general purpose, technique for faster sampling of nodes over an online social network. Specifically, unlike traditional random walks which wait for the convergence of sampling distribution to a predetermined target distribution - a waiting process that incurs a high query cost - we develop WALK-ESTIMATE, which starts with a much shorter random walk, and then proactively estimate the sampling probability for the node taken before using acceptance-rejection sampling to adjust the sampling probability to the predetermined target distribution. We present a novel backward random walk technique which provides provably unbiased estimations for the sampling probability, and demonstrate the superiority of WALK-ESTIMATE over traditional random walks through theoretical analysis and extensive experiments over real world online social networks.

Original languageEnglish (US)
Pages (from-to)678-689
Number of pages12
JournalProceedings of the VLDB Endowment
Volume8
Issue number6
DOIs
StatePublished - Jan 1 2015
Event41st International Conference on Very Large Data Bases, VLDB 2015 - Kohala Coast, United States
Duration: Aug 31 2015Sep 4 2015

All Science Journal Classification (ASJC) codes

  • Computer Science (miscellaneous)
  • Computer Science(all)

Fingerprint Dive into the research topics of 'Walk, not wait: Faster sampling over online social networks'. Together they form a unique fingerprint.

  • Cite this

    Nazi, A., Zhou, Z., Thirumuruganathan, S., Zhang, N., & Das, G. (2015). Walk, not wait: Faster sampling over online social networks. Proceedings of the VLDB Endowment, 8(6), 678-689. https://doi.org/10.14778/2735703.2735707