PrivyProbe

Inspiration

Privacy and personal data tracking has been a rising issue especially the given the rapid advancement of technology tools. The Electronic Frontier Foundation (EFF) advocates for user privacy and free expression, notably through tools like their browser Privacy Badger. My project, PrivyProbe aims to create a data analysis/visualization dashboard for Privacy Badger tracker data, facilitating improvements, insights, and increased public interest in privacy protection.

What it does

After uploading a json file from the Privacy Badger to PrivyProbe, a dashboard with numerous analytical charts is displayed, including statistics for number of sites tracked, number of trackers, a bar chart for top sites that track user data, a distribution of tracker types, and top tracked site categories. This can allow for users and the Privacy Badger team to better understand tracker activity and analyzing tracker trends over time.

How we built it

I used React, JavaScript, Node.js, and CSS to build the front-end of the project. I also utilized Ant Design for some component designs and ApexCharts to make the reactive graphics. The backend, utilizing Express.js and Multer middleware, allows file uploads via a POST request to '/upload', saving the uploaded file as 'data.json' and responding with it, while also handling cross-origin requests using CORS.

Challenges we ran into

It was very difficult to integrate an API and natural language processing model that was able to categorize a website both for free and in a quick manner. After hours of researching and trying, I found Klazify, which was able scan a site's information and determine what categories would best fit it. It was also challenging trying to find a way to turn the data that Klazify returned into a chart since it is locally stored on the user's instance and not saved as a file.

Accomplishments that we're proud of

I'm very proud of how the backend data was able to be converted into very beautifully displayed charts and how the layout of the dashboard is easy to understand. I'm also very proud that I was able to integrate a machine learning API for site categorization in such a short amount of time. I always enjoy seeing the front-end of the site coming together, and I feel that the eye-catching visuals were well put together.

What we learned

This was my first time working on a project in the field of data security and privacy, and I feel that I have really gained an interest in the topic after having done research to create this project. I have also never created charts in terms of web development before, so I learned a lot about making graphics from json data.