Connect with us

Hi, what are you looking for?

Tech & Science

Noiszy interview: How a browser plugin masks your digital tracks (Includes interview)

We originally featured Noiszy last month, after the plugin launched. Available for Chrome and soon to be officially launched for Firefox, Noiszy works by layering your browsing history with additional automatically-generated data.
Noiszy lives in its own browser tab and picks sites to visit from a list of approved sources. When it’s turned on, the software randomly navigates to a site and waits a minute on the page, simulating a human user’s behaviour.
When the timer expires, the plugin makes a die roll. It has a three quarters chance of moving to another page on the same site. In the other cases, it will change domain, picking a new site from the list.
Noiszy’s website claims this artificial “noise” can improve the climate of digital data tracking. It reduces the value of the information collected by internet providers and services. It’s designed to create change by making data less meaningful, reducing its ability to be manipulated by governments and corporations.
READ MORE: Microsoft’s new Windows 10 S looks to answer the Chromebook
Digital Journal spoke to Noiszy creator Angela Grammatas to learn more about the plugin and its design. The conversation is published below:
Digital Journal: Hi Angela. Thank you for participating in this discussion. Could you introduce yourself and tell us where the idea for Noiszy came from?
Angela Grammatas [Noiszy]: I’ve worked with data – specifically online site and marketing data – for around 15 years. In that time I’ve come to focus on implementation and controlling how data is collected. I know how to work with and filter data in most common analytics tools.
DJ: Where did the inspiration for Noiszy originate? What issue are you trying to solve here?
AG: The inspiration for Noiszy was a combination of two things: the “filter bubble” phenomenon, and the recent ISP data collection legislation. If a single site or ad platform collects my data and uses it to target and personalize my web experience, I can easily circumvent that by using different browsers or incognito windows.
If that’s done at the ISP level, however, there’s no way around it. I thought of Noiszy as a new way to burst that bubble, and at the same time, make the collected data less valuable to the collectors.
DJ: What do you mean when you talk about the “filter bubble?” How is this phenomenon impacting how we interpret digital content?
AG: A filter bubble is the reduction of a user’s perspective resulting from algorithmically-chosen, personalized online content. Basically, algorithms serve us content that we’re most likely to click on, based largely on past behaviour. You’re likely to see content similar to what you’ve already viewed, creating an echo chamber effect.
DJ: Your website explains how Noiszy operates based on a predetermined list of sites to visit, rather than randomly navigating to anywhere on the web. Why have you opted to use this approach? How is this more effective than randomly selecting sites?
AG: Great question! There are two main reasons here. Firstly, truly random data wouldn’t look real. It would be very easy to pick real behaviour out of a random cloud. Picking out real behaviour from a cloud of plausible data is much more difficult. Secondly, for Noiszy to have an effect on personalization and targeting algorithms, we have to send Noiszy data to sites that use those tools.
I focused on sites within a single vertical because they often share data and target users across sites. For example, I want to target scenarios such as Jane viewed an article about cats on A.com so she’s shown cat promotions on B.com.
I chose news sites because they use personalization and targeting algorithms to promote content, and because I’m concerned about the “filter bubble” and its effect on the way people perceive the world. I wanted to have an impact there.
DJ: The way your algorithm works seems very regular and looks like it could be predictable. Noiszy consistently navigates to a site, waits a minute and then moves on, with a one quarter chance of changing domain. How do you prevent ISPs from detecting the pattern, or is this a non-issue?
AG: For the areas Noiszy addresses, it’s (mostly) a non-issue. I designed behaviour that would be indistinguishable from “real” data by common web analytics, personalization, and targeting tools.
Those tools – for now at least! – generally don’t have the capability to filter out Noiszy traffic. Things like using the user’s own browser, having a random-but-believable number of pageviews on a given domain, and restricting onsite navigation to links that actually do exist on the page, make the traffic look completely plausible.
DJ: But technically it’s possible for Noiszy’s browsing to be detected?
AG: If this data is collected by ISPs, and they have the interest in filtering it out, then technically it would be possible to do so. But, since the patterns are plausible, filtering that out would likely mean filtering out many false positives as well.
DJ: What’s the biggest issue with Noiszy right now, in terms of the potential of being tracked?
The ~1 minute delay between pageviews – this will be updated in the near future for a more variable delay.
DJ: Noiszy is being launched at a time when cybersecurity and privacy is highly topical. How do you see it fitting into the current climate of tracking and monitoring tools?
AG: Everywhere we go online, we leave digital footprints behind. There are many tools to mitigate this, like Noisify for Facebook and Internet Noise for Google searches, but if you want to participate in the online world, it’s nearly impossible to entirely avoid leaving these tracks.
Much of the privacy conversation focuses on data collection, and while that’s important, I think it’s too simplistic. As users we should demand that creating data does not implicitly mean that it can be processed, sold, and used in ways that we don’t agree to.
DJ: So you want Noiszy to help start a new conversation around data collection, where the users can take control?
We need to talk more about data usage and what it means to develop and use algorithms responsibly. I think of Noiszy data as a way for people to raise their voices. It’s a form of resistance.
Author’s Note: Since this interview was completed, Noiszy has been updated with changes that address some of the potential issues mentioned in the article. The site list is now expanded to include the top 40 news sites, according to Alexa’s global ranking. Users can also add their own sites. Additionally, the delay between page changes is now randomised, making it harder to track Noiszy activity.
Digital Journal thanks Angela Grammatas for participating in this discussion.
You can download the Noiszy extension from the Chrome Web Store.
READ NEXT: Creating digital noise, our original report on Noiszy.

Written By

You may also like:

Business

Speaking at the Innovation Week YYC Launch Party, Naheed Nenshi shared his vision for how Alberta can harness its innovation potential.

Life

COVID-19 pandemic did not lead to an excess in suicide mortality in 2020, that does not mean that the pandemic wasn't related, in some...

Tech & Science

Video game "Astro Bot" -- a family-friendly sci-fi adventure -- was named 2024's Game of the Year on Thursday - Copyright AFP Philip FONGVideo...