Abstract | ||
---|---|---|
Within OSNs, many of our supposedly online friends may instead be fake accounts called social bots, part of large groups that purposely re-share targeted content. Here, we study retweeting behaviors on Twitter, with the ultimate goal of detecting retweeting social bots.We collect a dataset of 10M retweets. We design a novel visualization that we leverage to highlight benign and malicious patterns of retweeting activity. In this way, we uncover a ?normal" retweeting pattern that is peculiar of human-operated accounts, and suspicious patterns related to bot activities. Then, we propose a bot detection technique that stems from the previous exploration of retweeting behaviors. Our technique, called Retweet-Buster (RTbust), leverages unsupervised feature extraction and clustering. An LSTM autoencoder converts the retweet time series into compact and informative latent feature vectors, which are then clustered with a hierarchical density-based algorithm. Accounts belonging to large clusters characterized by malicious retweeting patterns are labeled as bots. RTbust obtains excellent detection results, with F1=0.87, whereas competitors achieve F1?0.76.Finally, we apply RTbust to a large dataset of retweets, uncovering 2 previously unknown active botnets with hundreds of accounts.
|
Year | DOI | Venue |
---|---|---|
2019 | 10.1145/3292522.3326015 | WebSci '19: 11th ACM Conference on Web Science
Boston
Massachusetts
USA
June, 2019 |
Field | DocType | Volume |
Data mining,Feature vector,Autoencoder,Botnet,Computer science,Visualization,Feature extraction,Artificial intelligence,Cluster analysis,Machine learning | Journal | abs/1902.04506 |
ISBN | Citations | PageRank |
978-1-4503-6202-3 | 2 | 0.40 |
References | Authors | |
0 | 5 |
Name | Order | Citations | PageRank |
---|---|---|---|
Michele Mazza | 1 | 2 | 0.40 |
Cresci, S. | 2 | 235 | 21.79 |
Marco Avvenuti | 3 | 267 | 24.14 |
Walter Quattrociocchi | 4 | 19 | 4.88 |
Maurizio Tesconi | 5 | 281 | 32.06 |