Evaluating the End-User Experience of Private Browsing Mode

Nowadays, all major web browsers have a private browsing mode. However, the mode’s benefits and limitations are not particularly understood. Through the use of survey studies, prior work has found that most users are either unaware of private browsing or do not use it. Further, those who do use private browsing generally have misconceptions about what protection it provides.

However, prior work has not investigated why users misunderstand the benefits and limitations of private browsing. In this work, we do so by designing and conducting a two-part user study with 20 demographically-diverse participants: (1) a qualitative, interview-based study to explore users’ mental models of private browsing and its security goals; (2) a participatory design study to investigate whether existing browser disclosures, the in-browser explanations of private browsing mode, communicate the security goals of private browsing to users. We asked our participants to critique the browser disclosures of three web browsers: Brave, Firefox, and Google Chrome, and then design new ones.

We find that most participants had incorrect mental models of private browsing, influencing their understanding and usage of private browsing mode. Further, we find that existing browser disclosures are not only vague, but also misleading. None of the three studied browser disclosures communicates or explains the primary security goal of private browsing. Drawing from the results of our user study, we distill a set of design recommendations that we encourage browser designers to implement and test, in order to design more effective browser disclosures.

SpeedReader: Reader Mode Made Fast and Private

Most popular web browsers include “reader modes” that improve the user experience by removing un-useful page elements. Reader modes reformat the page to hide elements that are not related to the page’s main content. Such page elements include site navigation, advertising related videos and images, and most JavaScript. The intended end result is that users can enjoy the content they are interested in, without distraction.

In this work, we consider whether the “reader mode” can be widened to also provide performance and privacy improvements. Instead of its use as a post-render feature to clean up the clutter on a page we propose SpeedReader as an alternative multistep pipeline that is part of the rendering pipeline. Once the tool decides during the initial phase of a page load that a page is suitable for reader mode use, it directly applies document tree translation before the page is rendered.

Based on our measurements, we believe that SpeedReader can be continuously enabled in order to drastically improve end-user experience, especially on slower mobile connections. Combined with our approach to predicting which pages should be rendered in reader mode with 91% accuracy, it achieves drastic speedups and bandwidth reductions of up to 27x and 84x respectively on average. We further find that our novel “reader mode” approach brings with it significant privacy improvements to users. Our approach effectively removes all commonly recognized trackers, issuing 115 fewer requests to third parties, and interacts with 64 fewer trackers on average, on transformed pages.

Who Filters the Filters: Understanding the Growth, Usefulness and Efficiency of Crowdsourced Ad Blocking

Ad and tracking blocking extensions are among the most popular browser extensions. These extensions typically rely on filter lists to decide whether a URL is associated with tracking or advertising. Millions of web users rely on these lists to protect their privacy and improve their browsing experience. Despite their importance, the growth and health of these filter lists is poorly understood. These lists are maintained by a small number of contributors, who use a variety of undocumented heuristics to determine what rules should be included. These lists quickly accumulate rules over time, and rules are rarely removed. As a result, users’ browsing experiences are degraded as the number of stale, dead or otherwise not useful rules increasingly dwarfs the number of useful rules, with no attenuating benefit. This paper improves the understanding of crowdsourced filter lists by studying EasyList, the most popular filter list. We find that, over its 9 year history, EasyList has grown from several hundred rules, to well over 60,000. We then apply EasyList to a sample of 10,000 websites, and find that 90.16% of the resource blocking rules in EasyList provide no benefit to users, in common browsing scenarios. Based on these results, we provide a taxonomy of the ways advertisers evade EasyList rules. Finally, we propose optimizations for popular ad-blocking tools that provide over 99% of the coverage of existing tools, but 62.5% faster.

AdGraph: A Machine Learning Approach to Automatic and Effective Adblocking

Filter lists are widely deployed by adblockers to block ads and other forms of undesirable content in web browsers. However, these filter lists are manually curated based on informal crowdsourced feedback, which brings with it a significant number of maintenance challenges. To address these challenges, we propose a machine learning approach for automatic and effective adblocking called AdGraph. Our approach relies on information obtained from multiple layers of the web stack (HTML, HTTP, and JavaScript) to train a machine learning classifier to block ads and trackers. Our evaluation on Alexa top-10K websites shows that AdGraph automatically and effectively blocks ads and trackers with 97.7% accuracy. Our manual analysis shows that AdGraph has better recall than filter lists, it blocks 16% more ads and trackers with 65% accuracy. We also show that AdGraph is fairly robust against adversarial obfuscation by publishers and advertisers that bypass filter lists.