MinerAlert: an hybrid approach for web mining detection

Abstract

The introduction of new memory-based crypto-mining techniques and the rise of new web technologies like WebAssembly, made the use of browsers for crypto-currencies mining more and more convenient and popular. That, in turn, originated a new form of computer piracy, called cryptojacking, which is rapidly gaining ground on the web. A cryptojacking site exploits its visitors’ hardware resources to secretly mine crypto-currencies. This paper analyzes current web-based cryptojacking detection methods in order to propose a novel hybrid strategy. Current detection methods are found to require either considerable computer administration skills or execution privileges usually not available to common users. In this view, a method, named MinerAlert, has been designed and proposed, aiming at detecting in real-time sites performing cryptojacking. To address the limitations of current methods, the method implementation has been achieved through a browser extension. The present paper describes the method’s details and its implementation. It also reports the experimental results of its utilization, showing its positive performances in terms of ease of use, successful detections and speed.