TY - CHAP

T1 - Comparing the Sensitivity of Social Networks, Web Graphs, and Random Graphs with Respect to Vertex Removal

AU - Martin, Christoph

AU - Niemeyer, Peter

N1 - Conference code: 11

PY - 2016/2/5

Y1 - 2016/2/5

N2 - The sensitivity of networks to the removal of vertices has been studied extensively over the last 15 years. A common approach to measuring this sensitivity is (i) successively removing vertices following a specific removal strategy and (ii) comparing the original and the modified network using a specific comparison method. In this paper we apply a wide range of removal strategies and comparison methods in order to study the sensitivity of medium-sized networks from the real world and randomly generated networks. In the first part of our study we observe that social networks and web graphs differ in sensitivity. When removing vertices, social networks are robust, web graphs are not. This effect is consistent with the work of Boldi et al. who analyzed very large social networks and web graphs. For randomly generated networks we find that their sensitivity depends significantly on the comparison method. The choice of removal strategy has surprisingly marginal impact on the sensitivity for removal strategies derived from common centrality measures. However, the removal strategy has a strong impact when removing the vertices in random order.

AB - The sensitivity of networks to the removal of vertices has been studied extensively over the last 15 years. A common approach to measuring this sensitivity is (i) successively removing vertices following a specific removal strategy and (ii) comparing the original and the modified network using a specific comparison method. In this paper we apply a wide range of removal strategies and comparison methods in order to study the sensitivity of medium-sized networks from the real world and randomly generated networks. In the first part of our study we observe that social networks and web graphs differ in sensitivity. When removing vertices, social networks are robust, web graphs are not. This effect is consistent with the work of Boldi et al. who analyzed very large social networks and web graphs. For randomly generated networks we find that their sensitivity depends significantly on the comparison method. The choice of removal strategy has surprisingly marginal impact on the sensitivity for removal strategies derived from common centrality measures. However, the removal strategy has a strong impact when removing the vertices in random order.

KW - Business informatics

KW - centrality measure

KW - complex networks

KW - random graphs

KW - robustness analysis

UR - http://www.scopus.com/inward/record.url?scp=84966341118&partnerID=8YFLogxK

U2 - 10.1109/SITIS.2015.22

DO - 10.1109/SITIS.2015.22

M3 - Article in conference proceedings

T3 - Proceedings - 11th International Conference on Signal-Image Technology and Internet-Based Systems, SITIS 2015

SP - 460

EP - 467

BT - 2015 11th International Conference on Signal-Image Technology & Internet-Based Systems (SITIS)

A2 - Yetongnon, Kokou

A2 - Dipanda, Albert

A2 - Chbeir, Richard

PB - IEEE - Institute of Electrical and Electronics Engineers Inc.

T2 - International Conference on Signal-Image Technology & Internet-Based Systems 2015

Y2 - 23 November 2015 through 27 November 2015

ER -