Physics > Physics and Society
[Submitted on 11 Sep 2015 (v1), revised 6 Mar 2018 (this version, v3), latest version 30 Jul 2018 (v5)]
Title:Local Determines Global: Identification and Quantification of Influential Spreaders in Large Scale Social Networks
View PDFAbstract:Measuring and optimizing the spread of influence in big-data online social networks are important for designing efficient strategies for false information surveillance and viral marketing. As the viral spread of information on social network is a global process, it is commonly believed that measuring the influence of nodes and optimizing viral spreading would require the whole network information. By mapping the spreading dynamics onto bond percolation in statistical physics, we find that in many stochastic spreading events, the information spreading happens in only one of two well separated phases: a locally confined phase with a small number of influenced nodes, and a global viral spreading phase with a fixed fraction of whole network nodes that is invariant with respect to initial spreaders and realizations. The global and local phases are clearly separated, which allows us to distinguish between them by a small characteristic infection size. Thus, without the need of information about the global network structure, any nodes' global influence on the entire network can be accurately measured based on this local characteristic infection size. This motivates an efficient algorithm with constant time complexity on the long standing problem of best seed spreaders selection, whose performance remarkably close to the true optimum.
Submission history
From: Shenggong Ji [view email][v1] Fri, 11 Sep 2015 12:51:10 UTC (4,969 KB)
[v2] Fri, 29 Jan 2016 14:42:17 UTC (4,656 KB)
[v3] Tue, 6 Mar 2018 13:49:16 UTC (3,069 KB)
[v4] Wed, 23 May 2018 00:23:48 UTC (2,975 KB)
[v5] Mon, 30 Jul 2018 05:15:12 UTC (2,975 KB)
Current browse context:
physics.soc-ph
References & Citations
Bibliographic and Citation Tools
Bibliographic Explorer (What is the Explorer?)
Litmaps (What is Litmaps?)
scite Smart Citations (What are Smart Citations?)
Code, Data and Media Associated with this Article
CatalyzeX Code Finder for Papers (What is CatalyzeX?)
DagsHub (What is DagsHub?)
Gotit.pub (What is GotitPub?)
Papers with Code (What is Papers with Code?)
ScienceCast (What is ScienceCast?)
Demos
Recommenders and Search Tools
Influence Flower (What are Influence Flowers?)
Connected Papers (What is Connected Papers?)
CORE Recommender (What is CORE?)
arXivLabs: experimental projects with community collaborators
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.