We further test our classifiers and several popular PP-Tools on a corpus ofĤK websites with 135K JS. When trained only on samples of tracking JS, our classifiersĪchieve an accuracy of 99%, where the best of the PP-Tools achieved an accuracy

One-class machine learning classifiers using syntactic and semantic featuresĮxtracted from JS. Web developers often borrow and customize existing pieces of code in order toĮmbed tracking (resp. That tracking JS share structural similarities that can be used toĭifferentiate them from functional JS. To improve this balance, we examine the two classes of JS and hypothesize To assess the performance of current web PP-Tools. To the best of our knowledge, this is the first study Ineffective in finding the right balance between blocking tracking JS andĪllowing functional JS. Labelled dataset composed of 2,612 JS, we show how current PP-Tools are Website from tracking JS would thus be useful. A mechanism that can isolate JS necessary for proper functioning of the

Tracking, often executed via JS, may suffer from poor performance and limited Privacy preserving tools (PP-Tools) that have been developed to limit online Such blocking also affectsįunctionality of webpages and impairs user experience. Popular JavaScript programs (JS) in Web browsers.

Download a PDF of the paper titled Towards Seamless Tracking-Free Web: Improved Detection of Trackers via One-class Learning, by Muhammad Ikram and 4 other authors Download PDF Abstract: Numerous tools have been developed to aggressively block the execution of

Seamless web