ACSPRI Conferences, ACSPRI Social Science Methodology Conference 2014

Font Size:  Small  Medium  Large

Determining brokerage and closure processes from time-ordered interactions

Lucia Falzon, John M Dunn

Building: Holme Building
Room: MacCallum Room
Date: 2014-12-09 01:30 PM – 03:00 PM
Last modified: 2014-10-31


Social network analysis is widely used in the exploration of social relations and groups. It provides a general framework for exploring interactions, and allows regularities to be detected, described and compared across observable networks. The underlying assumption is that social activities can be understood, in part, through the regularities that characterise relationships between social actors. The investigation of static relations and their formation into substructures (such as triads) dominate the social network analysis literature. Brokerage and closure processes (e.g. transitive and cyclic closures) are typically treated as “atemporal” processes. In this paper we argue that the analysis of social interactions must consider their temporal nature. In particular, the exact sequence of communications and transactions trace trajectories through the network of interest that enable us to study the diffusion of information and the potential brokerage roles played by the actors involved.
The ever-increasing use of the Internet, through services such as email and social networking, has led to the production of time-stamped data on information-sharing activities that enable us to re-construct the unfolding of social processes over time. We present newly developed techniques for the analysis of time-ordered interactions that measure explicitly the duration and path-lengths of specific sequences of exchanges. Importantly, these techniques enable us to characterize the temporal behaviour of such processes by investigating their completion time and the brokerage roles involved in candidate interaction patterns.