Identifying suspicious URLs: an application of large-scale online learning

By Justin Ma, Lawrence K. Saul, Stefan Savage, Geoffrey M. Voelker. 2009

609 citations recorded by the source corpus.

Subfields: Computer Networks and Communications, Information Systems, Safety, Robustness & Interpretability, Signal Processing

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Networks and Communications

In Information Systems

In Safety, Robustness & Interpretability

In Signal Processing

Citations in the corpus

Improved asynchronous parallel optimization analysis for stochastic incremental methods (2018, methodology): “This data set contains more than 2 million URLs obtained at random from Yahoo’s directory listing (for the “benign” URLs) and from a large Web mail provider (for the “malicious” URLs).”
Learning privately from multiparty data (2016, background): “The Malicious URL Dataset (Ma et al., 2009) is a collection of examples of malicious URLs from a large Web mail provider.”
Perturbed Iterate Analysis for Asynchronous Stochastic Optimization (2015, neutral): “Logistic regression Synthetic 3M 10K 20 rcv1 [31] from [32] ≈ 700K ≈ 42K ≈ 73 url [33] from [32] ≈ 3.”
ASAGA: Asynchronous Parallel SAGA (2016, methodology): “(2011) and re-used in all the later papers that we mentioned in the related work section, with the notable exceptions of Mania et al. (2015) and Duchi et al.”
Beyond blacklists: learning to detect malicious web sites from suspicious URLs (2009, methodology): “We address the issue in subsequent work by using online learning algorithms [14].”
Fully Distributed Privacy Preserving Mini-batch Gradient Descent Learning (2015, background): “provide a nice summary of online learning for large scale data [15].”
Cyclades: Conflict-free Asynchronous Machine Learning (2016, methodology): “Specifically, we run classification for the URL dataset [15] contains ⇠ 2.”
PhishZoo: Detecting Phishing Websites by Looking at Them (2011, background): “Host features of the URL include properties of IP address, the owner of the site, DNS properties such as TTL, and geographical location [10].”
Malicious URL Detection using Machine Learning: A Survey (2017, background): “For the 2 million URLs, [28] obtained over a million host-based features.”
Loda: Lightweight on-line detector of anomalies (2016, methodology): “Finally, Loda's capability to efficiently process big datasets is demonstrated on the WEB UCI dataset (Ma et al. 2009) with 2.4 million samples and 3.2 million features.”
Learning to Detect Phishing Webpages (2014, methodology): “have been shown to perform better compared to some batch algorithms when trained continuously with variable-feature sets in the context of detecting malicious URLs [31].”
AUC Maximization in the Era of Big Data and AI: A Survey (2022, background): “Such datasets are abundant in online spam filtering [145], ad click prediction [112], and identifying malicious URLs [108].”
Proactive Discovery of Phishing Related Domain Names (2012, methodology): “…on the ﬂy, domains as malicious or benign is widely used, with either batch methods such as SVM, Naive Bayes, Logistic Regression(like in [2–4]), or on-line classiﬁ-cation algorithms such as Conﬁdence Weighted (CW), Adaptive Regularization of Weight (AROW), Passive-Aggressive (PA) (see [5, 9, 15]).”