Proceedings of the 14th ACM International Conference on Web Search and Data Mining 2021
DOI: 10.1145/3437963.3441806
|View full text |Cite
|
Sign up to set email alerts
|

F-FADE: Frequency Factorization for Anomaly Detection in Edge Streams

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
3
2

Citation Types

0
23
0

Year Published

2021
2021
2024
2024

Publication Types

Select...
3
3
2

Relationship

0
8

Authors

Journals

citations
Cited by 36 publications
(23 citation statements)
references
References 25 publications
0
23
0
Order By: Relevance
“…PENminer [8] explores the persistence of activity snippets, i.e., the length and regularity of edge-update sequences' reoccurrences. F-FADE [15] aims to detect anomalous interaction patterns by factorizing the frequency of those patterns. These methods can effectively detect periodic patterns, but they require a considerable amount of time to explore and maintain possible patterns in the network.…”
Section: Related Workmentioning
confidence: 99%
See 1 more Smart Citation
“…PENminer [8] explores the persistence of activity snippets, i.e., the length and regularity of edge-update sequences' reoccurrences. F-FADE [15] aims to detect anomalous interaction patterns by factorizing the frequency of those patterns. These methods can effectively detect periodic patterns, but they require a considerable amount of time to explore and maintain possible patterns in the network.…”
Section: Related Workmentioning
confidence: 99%
“…On the other hand, there is another category of algorithms that aims to detect anomalies using patterns or motifs [8,15,43,38,30,33,55,3,23,29,41,40,35,31,7,20,11]. However, many of these methods require active exploration of patterns or snippets, increasing memory and time requirements.…”
Section: Introductionmentioning
confidence: 99%
“…Though the variants of Isconna demonstrates a comparatively better accuracy than the state-of-the-art, however it suffers from parameter tuning. Very recently, Chang et al have proposed a frequency factorization approach (termed as F-fade) for anomaly detection in edge streams [5].…”
Section: Related Workmentioning
confidence: 99%
“…MIDAS [15] proposes a probabilistic framework to detect microcluster anomalies (suddenly arriving of groups of suspiciously similar edges). F-FADE [13] uses a frequency-factorization technique to model the distributions of edge frequencies and flags edges with low likelihood of the observed frequency as anomaly.…”
Section: Related Workmentioning
confidence: 99%
“…More recently, [11], [12], [13] have proposed to detect individually surprising edges in the edge stream which have superior memory and time efficiency. However, due to the non-i.i.d.…”
Section: Introductionmentioning
confidence: 99%