DMCA
Sparsification of Influence Networks
| Citations: | 26 - 6 self |
BibTeX
@MISC{Mathioudakis_sparsificationof,
author = {Michael Mathioudakis and Francesco Bonchi and Carlos Castillo and Aristides Gionis and Antti Ukkonen},
title = {Sparsification of Influence Networks},
year = {}
}
Share
 |
 |
 |
 |
||
OpenURL
Abstract
We present Spine, an efficient algorithm for finding the “backbone ” of an influence network. Given a social graph and a log of past propagations, we build an instance of the independent-cascade model that describes the propagations. We aim at reducing the complexity of that model, while preserving most of its accuracy in describing the data. We show that the problem is inapproximable and we presentanoptimal, dynamic-programmingalgorithm,whose search space, albeit exponential, is typically much smaller than that of the brute force, exhaustive-search approach. Seeking a practical, scalable approach to sparsification, we devise Spine, a greedy, efficient algorithm with practically little compromise in quality. We claim that sparsification is a fundamental datareduction operation with many applications, ranging from visualization to exploratory and descriptive data analysis. As a proof of concept, we use Spine on real-world datasets, revealing the backbone of their influence-propagation networks. Moreover, we apply Spine as a pre-processing step for the influence-maximization problem, showing that computations on sparsified models give up little accuracy, but yield significant improvements in terms of scalability.
Keyphrases
influence network efficient algorithm real-world datasets search space little accuracy past propagation many application exhaustive-search approach scalable approach influence-maximization problem yield significant improvement descriptive data analysis brute force little compromise present spine independent-cascade model fundamental datareduction operation influence-propagation network sparsified model pre-processing step social graph
