Knowledge Graph Metadata: What Facebook really knows about you?

If you’re one of the holdouts who still has a Facebook account you’ll be happy to know that Big Zuck now lets you control — and delete — more of your personal data. In fact, the social network will now show you the ads you clicked, information about the skin tone sensors it uses to identify you online, and many other important pieces of information you didn’t even know it was collecting.

But how can you get a sense of this massive trove of information?

Without proper analytics and visualization tools, it’s useless.

Knowledge Graph

Over the last decade tech giants like Amazon, Facebook, and Google invested millions to create their own knowledge graphs that power their search engines and allow them to understand the context of a query and a meaning behind specific user intent.

I decided to flip the script and build a knowledge graph out of the data Facebook collected about me to learn a bit more about the state of the ad industry and, dare I say, myself.

I was able to do this thanks to Knowledge Graph Base, the latest product from Thinknum, which allows you to build knowledge graphs without writing a line of code.

With a quick JSON file integration setup, I was able to understand what bots thought about me. Here are a few things I found after running some simple queries.

38% of advertisers targeting me on Facebook are car dealerships

As someone who has never owned a car and is not planning to buy one, that came as a surprise.

What was even more interesting to me was the range of car models and locations where I was targeted. Honda Dealers in Oklahoma and Dahmer Buick of Kansas City targeted me at the same time as BMW of Denver. I have no connection to any of these places.

Is Facebook’s algorithm clueless or is the data they share skewed? You decide.

Interestingly, a few of my co-workers also suffered from long distance car sales targeting syndrome, which suggests something else is triggering these ridiculous ads.

The ads I interacted with had minimum correlation with those targeting me.

Looking into the list of advertisers who I interacted with on Facebook over time I learned that fashion ads are my favorite, but that none of the companies that I interacted with ever added me to their marketing lists.

In other words, the companies I care about don’t care about me.

Insert sad face emoji here.

Cardi B is targeting me on Facebook!

Wow! Belcalis Marlenis Almánzar is totally into me! Finally an example of correct targeting.

As most datasets are siloed producing an information overload challenge, graphs bases are important tools to sparks new ideas and find unexpected connections.

Thinknum’s KgBase, or Knowledge Graph Base, is a collaborative, robust database with versioning, analytics, and visualizations. It’s easy to use and I encourage you to try it.

With KgBase, any community or individual can create knowledge graphs to build insights about their data. Import your CSVs and spreadsheets, or use our API to work on data together.

The goal? For all of us to understand just what Facebook — and other powerful companies — think about us.

Learn how to use Facebook integration via Kgbase with your own data here.