2010
DOI: 10.1007/978-3-642-13672-6_40
|View full text |Cite
|
Sign up to set email alerts
|

oddball: Spotting Anomalies in Weighted Graphs

Abstract: Given a large, weighted graph, how can we find anomalies? Which rules should be violated, before we label a node as an anomaly? We propose the OddBall algorithm, to find such nodes. The contributions are the following: (a) we discover several new rules (power laws) in density, weights, ranks and eigenvalues that seem to govern the socalled "neighborhood sub-graphs" and we show how to use these rules for anomaly detection; (b) we carefully choose features, and design OddBall, so that it is scalable and it can w… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
3
1
1

Citation Types

0
372
0

Year Published

2011
2011
2016
2016

Publication Types

Select...
4
3
2

Relationship

2
7

Authors

Journals

citations
Cited by 422 publications
(372 citation statements)
references
References 23 publications
0
372
0
Order By: Relevance
“…The algorithms rely on extracting graph-level (global) features and tracking these metrics over time. In [17], the authors propose to detect abnormal nodes from weighted graphs based on features and patterns from the egonet. There has also been work on using local and global structural features to improve the performance of network classifiers [18].…”
Section: Related Workmentioning
confidence: 99%
“…The algorithms rely on extracting graph-level (global) features and tracking these metrics over time. In [17], the authors propose to detect abnormal nodes from weighted graphs based on features and patterns from the egonet. There has also been work on using local and global structural features to improve the performance of network classifiers [18].…”
Section: Related Workmentioning
confidence: 99%
“…• at which nodes or subgraphs to start the exploration [67], [57]? • if a user has a particular goal in mind, can it be predicted by their interactions?…”
Section: B Global and Local Viewsmentioning
confidence: 99%
“…Akoglu et al [3] proposed a method to assign anomaly scores to nodes based on egonet properties in weighted networks. Our framework allows us to incorporate such properties as application-specific filters.…”
Section: Related Workmentioning
confidence: 99%