PharmaGuard: Automatic identification of illegal search-indexed online pharmacies

Abstract

We present PharmaGuard, a novel system for the automatic discovery of illegal online pharmacies, aimed at assisting law-enforcement toward their early identification, blacklisting and shutdown. Given a previously labelled set of examples, the system is able to learn a profile of (illegal) pharmacies, and then exploit it to discover never-before-seen instances indexed by popular web search engines. Our experiments, performed on webpages found in the wild, indicate that our approach is lightweight, allows for high accuracy and can substantially complement state-of-the-art blacklists. We also present a report on the detected online pharmacies that better highlights the relevance of this threat for Internet users.