Leveraging node neighborhoods and egograph topology for better bot detection in social graphs

Due to their popularity, online social networks are a popular target for spam, scams, malware distribution and more recently state-actor propaganda. In this paper, we review a number of recent approaches to fake account and bot classification. Based on this review and our experiments, we propose our...

Full description

Saved in:
Bibliographic Details
Published in:Social network analysis and mining 2021-12, Vol.11 (1), p.10, Article 10
Main Authors: Bebensee, Björn, Nazarov, Nagmat, Zhang, Byoung-Tak
Format: Article
Language:English
Subjects:
Applications of Graph Theory and Complex Networks
Classification
Computer Science
Data Mining and Knowledge Discovery
Economics
Ethnicity
Experiments
Game Theory
Graphs
Humanities
Law
Metadata
Methodology of the Social Sciences
Neighborhoods
Network topologies
Original Article
Popularity
Propaganda
Public opinion
Semantics
Social and Behav. Sciences
Social networks
Statistics for Social Sciences
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Due to their popularity, online social networks are a popular target for spam, scams, malware distribution and more recently state-actor propaganda. In this paper, we review a number of recent approaches to fake account and bot classification. Based on this review and our experiments, we propose our own method which leverages the social graph’s topology and differences in egographs of legitimate and fake user accounts to improve identification of the latter. We evaluate our approach against other common approaches on a real-world dataset of users of the social network Twitter.
ISSN:1869-5450
1869-5469
DOI:10.1007/s13278-020-00713-z