RUBINET Data Sets

Online Social Networks: Anonymized Data from Third-Party Facebook Applications

  • Dataset #1: single tar file with README
    Description: This data set includes a 3-week long trace (ending April 1, 2008) of activity on three Facebook-based applications (Fighters' Club, Got Love?, and Hugged). A comprehensive analysis of this data set has been published in IMC'08, and this data set is made available to the internet measurement community for further exploration of activity originating from third party social networking applications.
    Paper:
    • A. Nazir, S. Raza, and C-N. Chuah, Unveiling Facebook: A Measurement Study of Social Network Based Applications, ACM Internet Measurement Conference (IMC), October 2008. [pdf]

  • Dataset #2: single tar file (1.5GB+) with README
    Description:This data set includes experimentation results from PlanetLab nodes as well as passive measurement performed at the third-party application servers for six popular Facebook applications, as used in the IMC'09 publication titled 'Network Level Footprints of Facebook Applications.' This data set is made available to the internet measurement community.
    Paper:
    • A. Nazir, S. Raza, D. Gupta, C-N. Chuah, B. Krishnamurthy, "Network Level Footprints of Facebook Applications," ACM/USENIX Internet Measurement Conference (IMC), November 2009. [pdf]

  • Dataset #3: iHeart, Hugged, and iSmile
    Description: This data set includes anonymized user activity data on three Facebook-based gifting applications (iHeart, Hugged, and ismile). For each of the application, the data set contains the time stamp of individual user activity, e.g., sending of a gift item from user X to Y.
    • Hugged was launched in February 2008 and installed by 28 million users on Facebook by August 2010. Users share graphics of hugging characters with their Facebook friends. Our user activity traces capture ~1.6 billion user activities spanning 68 weeks that include the peak usage period of the application.
    • iSmile was launched in August 2008, and installed by 43 million Facebook users by August 2010. Users share graphics of smiling characters with their Facebook friends. Our trace capture user activities spanning 124 weeks.
    • iHeart was launched in June 2009, and grew to 30 million monthly active users by December 2009, and was installed by 76 million users by August 2010. Users share decorative heart-shaped graphics with their Facebook friends. Our trace captured 2.2 billion user activities spanning 58 weeks, covering the most of the lifespan of the application (including initial launch, pre-peak, peak, and post-peak periods).
    Paper:
    • M. R. Rahman, J. Han, and C-N. Chuah, "Unveiling the Adoption and Cascading Process of OSN-based Gifting Applications," IEEE INFOCOM, April 2015. [pdf]
    • M. R. Rahman, P-A. Noel, C-N. Chuah, B. Krishnamurthy, and R. M. D'Souza, "Peeking into Invitation-based Adoption Process of OSN-based Applications," ACM Computer Communication Review (CCR), vol. 44, no. 1, pp. 21-27, January 2014. (Best of CCR, to be presented at ACM SIGCOMM 2014) [pdf]
    • H. Liu, A. Nazir, J. Joung, and C-N. Chuah, "Modeling/ Predicting the Evolution of User Activity Graphs on OSN-based Applications," WWW, May 2013. [pdf]
    • A. Nazir, A. Waagen, V. S. Vijayaraghavan, C-N. Chuah, R. D'Souza, and B. Krishnamurthy, "Beyond Friendship: Modeling User Activity Graphs on Social Network Based Gifting Applications" ACM Internet Measurement Conference, November, 2012. [pdf]

Bluetooth Connectivity Trace

  • Dataset: UC Davis Haggle - Unitrans Mobility Trace
    Description: This data set includes several traces about the available Bluetooth connectivity during a typical day on the Unitrans bus system at University of California, Davis. The data has been uploaded to the CRAWDAD. The data was collected using Intel iMotes, which contain an ARM core as well as a Bluetooth radio stack. The motes run a simple Bluetooth inquiry program. Periodocally (e.g., every 2 minutes) the device does an inquiry for a duration (e.g., 5 seconds). All devices detected during this inquiry period are noted. These experiments give a list of contacts - with start and end times - that a particular mote has during its lifetime. For each experiment, the polling interval was 2 minutes, and the inquiry duration was 5 seconds.
    Paper: