Social Scrubber

Inspiration

We’ve all heard the horror stories: one day, some Instagram influencer or musical artist posts a picture of themselves on social media. The next day, a deranged fan stalks them, and their encounter makes the headlines. Or maybe, more innocuously, you’ve seen one of your acquaintances accidentally give away some personally identifiable information in a photo or text, like their phone number or their address. The ubiquitous reach of social media makes security and privacy a precious commodity in the modern day. Our application, Social Scrubber, aims to help people filter potential identifying information out of their social media posts.

What it does

Our application, Social Scrubber, identifies which parts of a Facebook post may contain personally identifiable information (PII). Users install our application as a Chrome Extension which detects when the user is making a Facebook post. The application automatically identifies potential PII in the process of making the post, both in the text box as well as within the uploaded photos. When it identifies potential PII, it warns the user via a pop-up containing the offending information.

How I built it

We used the Azure Content Moderation API for text that identifies if a particular piece of text contains personally identifiable information (PII). We used Azure's OCR technology to find text and then scanned that text for PII. The API is hosted on Azure Web Services as part of an Express.js webapp.

Challenges I ran into

We ran into issues trying to retrieve images from the Facebook post because Facebook automatically compresses and crops the image into a thumbnail before the user posts. In particular, PII in the image was sometimes cut-off or blurry, which made it difficult for Azure to identify. We attempted to address this problem in various ways, including retrieving the original image from its source, trying to obtain the React component containing the image, and automatic resizing.

Accomplishments that I'm proud of

We are proud of the potential impact that this application could have on people’s lives. Revealing information like your SSN or your address on the internet can have drastic consequences. The aftermath of an easily preventable mistake such as not noticing something in the background of a photo can be life-altering.

What I learned

We learned how to host a website on Azure Web Services as well as how to use Azure’s Content Moderation and OCR APIs.

What's next for Social Scrubber

We would expand this extension to other social media platforms. We would also like to create comprehensive personal profiles based solely on information available on individual’s social media accounts.