Answers are grounded in this paper and cite their sources.
Typosquatting is a deceptive practice that involves exploiting typographical errors made by users when manually entering URLs. Although well-known and widely exploited, effective user-side mitigation tools for typosquatting remain scarce. In this work we propose an effective, user-side, typosquatting detection scheme called TypoDS , which combines and improves state-of-the-art typosquatting detection techniques and which is made freely available as a dedicated extension, named TypoAlert, for the popular web browser Chrome. TypoAlert is designed to automatically detect in real-time when a user visits a typosquatted domain and alert the user accordingly. The proposed approach relies on extracting informative features from URLs. These features are converted into quantitative scores, which are then combined by means of a threshold-based scoring function, whose threshold is experimentally determined, to classify a URL as legitimate or typosquatted. The proposed scheme has been validated on a dataset with a substantial number of labeled domains, and its performance has been compared against machine-learning-based classification algorithms. The results show that a TypoDS implementation is able to reach an accuracy of 98.41%.
No takes yet. Share an insight, caveat, or question.
Blefari et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: