Undermining and Strengthening Social Networks through Network Modification

TL;DR

This paper presents a flexible framework for network modification that can optimize various interventions to influence social networks, demonstrated through case studies on terrorist networks and entrepreneurial ecosystems.

Contribution

It introduces a comprehensive framework for network interventions that accounts for different modification types and recovery mechanisms, expanding the scope of network influence strategies.

Findings

Optimal node removal reduces terrorist attacks in simulations.

Strengthening ties enhances entrepreneurial ecosystems in developing countries.

Empirical calibration improves intervention effectiveness.

Abstract

Social networks have well documented effects at the individual and aggregate level. Consequently it is often useful to understand how an attempt to influence a network will change its structure and consequently achieve other goals. We develop a framework for network modification that allows for arbitrary objective functions, types of modification (e.g. edge weight addition, edge weight removal, node removal, and covariate value change), and recovery mechanisms (i.e. how a network responds to interventions). The framework outlined in this paper helps both to situate the existing work on network interventions but also opens up many new possibilities for intervening in networks. In particular use two case studies to highlight the potential impact of empirically calibrating the objective function and network recovery mechanisms as well as showing how interventions beyond node removal can be optimised. First, we simulate an optimal removal of nodes from the Noordin terrorist network in order to reduce the expected number of attacks (based on empirically predicting the terrorist collaboration network from multiple types of network ties). Second, we simulate optimally strengthening ties within entrepreneurial ecosystems in six developing countries. In both cases we estimate ERGM models to simulate how a network will endogenously evolve after intervention.