Outcome

Intention

In my beginning investigations for Data Decoder, I was looking at the various ways that identity is perceived on the internet. My initial think piece discussed identity through what it means to be anonymous in a public sphere like the internet, and how identity might be constructed (or reconstructed) through constantly changing means. Post review and class discussion I decided to take my investigation towards machine perception, or rather the ways in which our identity is being perceived on the internet by the companies & platforms for whom we create personal and identifiable data. My intention with Data Decoder is to help users retain some agency to the data they have created for these platforms by viewing with a tool that interprets the data and reveals it in an environment separate from the one in which it was created. The current artifact is a mockup of this data browser/interpretation tool, and the idea is that users would be able to view and interact with the data from these platforms in a more manageable way than what is given in response to a request for data from these platforms.

Outcome

Link to video

Link to figma prototype

Data decoder is a web app tool that makes it easy for users to view the data they have generated for technology platforms and services like Facebook, Amazon, Twitter, & Google. These online platforms, and many other ones, make ‘all’ (not actually) of the data that you own on their platforms downloadable following a request. Using Data Decoder, users can upload the data they’ve downloaded and the web app would parse the file types and make them more manageable to view. Users can get some insight into the various types of hidden data that is stored by these platforms about them, but also look at meta-statistics of all the data they’ve created over their life on the platforms.

Process

Initially, I had a direction of finding ways to subvert some of these forms of machine perception and algorithmic data collection and began with looking at existing projects and tools in this space. My first foray turned up projects like Unique Machine, Cover Your Tracks, and Am I Unique, which reveal the data collection that happens to uniquely identify someone online. Data Selfie (RIP) is a project that is most similar to Data Decoder in that it provides users a window into how their online presence is being perceived, but it relies on regenerating that data through tracking someone’s activity rather than using data provided by the platforms. With the aim to subvert, I spent time googling the ways in which someone could request a copy of their data from some of the big companies. Some of my requests were more successful than others, with a disclaimer from most that it could take up to 30 days to receive the data after my request. Browsing some of the data I received in the form of JSON, HTML, and CSV files, spurred the idea of a tool that would make it easier to interpret this. Funnily enough, the initial thought was, “My mom wouldn’t know how to open these in VS Code.” From there, it was a matter of conceiving of a basic mockup of what a Data Decoder web app might look like.

Reflection

I found this line of inquiry meaningful in that it brings up questions in the themes of privacy, quantified self, and machine perception, and further adds to the ongoing discourse about data collection from these large companies. What I found most interesting was the other student’s projects dealing with the same sort of issues, but manifesting in different ways of visualizing and interacting with the same data. Given more time, I would be curious to dive deeper into visualizing some of the data provided back by the companies, as well as find ways of maybe scoring the data that is actually given and the companies usability for requesting that data. In terms of feasibility, I think this project is definitely something that could be built, given the relative simplicity of the files produced by the requests for data.