Title: Variable social vector clocks for exploring user interactions in social communication networks

Authors: Ta-Yuan Hsu; Ajay D. Kshemkalyani

Addresses: Department of Electrical and Computer Engineering, University of Illinois at Chicago, Chicago, IL 60607, USA ' Department of Computer Science, University of Illinois at Chicago, Chicago, IL 60607, USA

Abstract: Social network communication analysis has drawn widespread attention in recent years. Vector clocks can be applied to capture the most recent communication with all other local peers in a social network. A modification of conventional social vector clocks has been previously proposed to deal with the issue of poor scalability without keeping whole temporal views. In this paper, our proposed framework maintains the low bound of how out-of-date each peer can be with respect to others, and also considers the shortest friendship separation to restrict how far information may be transmitted along time-respecting paths. To quantitatively analyse the influence of user interactions over different limitations of friendship distance, we also provide an adaptive incremental updating approach that can exactly recover the real situation in a specified upper bound of friendship distance. Experimental results also show that social vector clocks can be efficiently exploited to improve memory space requirements.

Keywords: vector clocks; social networks; n-degree separation; Twitter; social influence; space-based computing; user interactions; social network analysis; SNA; memory space.

DOI: 10.1504/IJSSC.2015.067997

International Journal of Space-Based and Situated Computing, 2015 Vol.5 No.1, pp.39 - 52

Received: 01 Aug 2014
Accepted: 18 Nov 2014

Published online: 31 Mar 2015 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article