Defending against large-scale crawls in online social networks

Defending against large-scale crawls in online social networks

Cached

Download Links

by Mainack Mondal , Krishna Gummadi , Bimal Viswanath , Allen Clement , Alan Mislove , Peter Druschel , Ansley Post

Citations:

1 - 1 self

Versions

Version 0

Version History

Metadata Version 0

Datum	Value	Source
TITLE	Defending against large-scale crawls in online social networks	SVM HeaderParse 0.2
AUTHOR NAME	Mainack Mondal	SVM HeaderParse 0.2
AUTHOR AFFIL	MPI-SWS	SVM HeaderParse 0.2
AUTHOR NAME	Krishna Gummadi	SVM HeaderParse 0.2
AUTHOR AFFIL	MPI-SWS	SVM HeaderParse 0.2
AUTHOR NAME	Bimal Viswanath	SVM HeaderParse 0.2
AUTHOR AFFIL	MPI-SWS	SVM HeaderParse 0.2
AUTHOR NAME	Allen Clement	SVM HeaderParse 0.2
AUTHOR AFFIL	MPI-SWS	SVM HeaderParse 0.2
AUTHOR NAME	Alan Mislove	SVM HeaderParse 0.2
AUTHOR AFFIL	Northeastern University	SVM HeaderParse 0.2
AUTHOR NAME	Peter Druschel	SVM HeaderParse 0.2
AUTHOR AFFIL	MPI-SWS	SVM HeaderParse 0.2
AUTHOR NAME	Ansley Post	SVM HeaderParse 0.2
ABSTRACT	Thwarting large-scale crawls of online social networks (OSNs) like Facebook and Renren is in the interest of both the users and the operators of these sites. OSN users wish to maintain control over their personal information, and OSN operators wish to protect their business assets and reputation. Traditional defenses against crawlers involve rate-limiting the browsing activity of individual users. However, these schemes are ineffective against crawlers with many accounts, be they fake accounts (Sybils) created by a crawler or compromised accounts of real users obtained on the black market. We propose Genie, a system that can be deployed by OSN operators to defend against crawlers. Genie’s design is based on the observation that the browsing patterns of normal users and crawlers are very different: most normal users limit their profile views to a small number of other users that are in their close network neighborhood, while even a crawler with many accounts needs to view profiles of users that are relatively more distant from the closest user account he controls. Genie exploits this fact by limiting the rate of profile views based on the connectivity and distance between a profile viewer and viewee in the social network. An experimental evaluation using real-world data gathered from a popular OSN shows that Genie frustrates large-scale crawling. Most browsing by ordinary users is not affected; the few users who are affected can recover easily by adding a few friend links. 1	SVM HeaderParse 0.2
CITATIONS	14 found	ParsCit 1.0