Social networks have been instrumental in spreading rumor such as fake news and false rumors. Research in rumor intervention to date has concentrated on launching an intervening campaign to limit the number of infectees. However, many emerging and important tasks focus more on early intervention. Social and psychological studies have revealed that rumors might evolve 70% of its original content after 6 transmissions. Therefore, ignoring earliness of intervention makes the intervening campaign downgrade rapidly due to the evolved content. In real social networks, the number of social actors is usually large, while the budget for an intervening campaign is relatively small. The limited budget makes early intervention particularly challenging. Nonetheless, we present an eicient containment method that promptly terminates the difusion with least cost. To our knowledge, this work is the irst to study the earliness of rumor intervention in a large real-world social network. Evaluations on a network of 3 million users show that the key social actors who earliest terminate the spread are not necessarily the most inluential users or friends of rumor initiators, and the proposed method efectively reduces the life span of rumors.

Original languageEnglish (US)
Title of host publicationWebSci 2019 - Proceedings of the 11th ACM Conference on Web Science
PublisherAssociation for Computing Machinery, Inc
Number of pages9
ISBN (Electronic)9781450362023
StatePublished - Jun 26 2019
Event11th ACM Conference on Web Science, WebSci 2019 - Boston, United States
Duration: Jun 30 2019Jul 3 2019

Publication series

NameWebSci 2019 - Proceedings of the 11th ACM Conference on Web Science


Conference11th ACM Conference on Web Science, WebSci 2019
Country/TerritoryUnited States


  • Classiication
  • Graph Mining
  • Social Media Mining
  • Social Network Analysis

ASJC Scopus subject areas

  • Computer Networks and Communications


Dive into the research topics of 'Debunking rumors in social networks: A timely approach'. Together they form a unique fingerprint.

Cite this