Tuesday, December 10, 2024

The Trouble with Facebook’s Graph Search

Datamation content and product recommendations are editorially independent. We may make money when you click on links to our partners. Learn More.

Facebook CEO Mark Zuckerberg announced a new feature yesterday called Graph Search.

Graph Search magnifies everything good and bad about Facebook. It’s Facebook on steroids.

To the extent that Facebook is good for connecting with family and friends, Graph Search makes Facebook great for that purpose. To the extent that Facebook is an invasion of privacy, a playground for stalkers, snoops, spammers and scammers, Graph Search makes Facebook great for those purposes also.

If you love Facebook, you’re going to love Graph Search. If you hate Facebook, you’re going to hate Graph Search.

Instead of a regular search engine, Graph Search is designed to simulate natural-language queries. Instead of typing keywords, you might write: “My friends who like football.” Graph Search is supposed to give you a list of everyone in your social graph who has ever clicked “Like” on anything football related, or expressed football enthusiasm in their status updates.

Essentially, Graph Search lets you build pseudo natural-language queries by mixing and matching people, places and interests. To combine all three, you might type “friends (people) in New York (place) who like baseball (interest)”. Or you can throw in “pictures” as a variable: “Pictures of friends of friends taken in New York.” Instagram photos and data will be included in the search results.

Once your results appear, you’ll also see something called the Power Bar, which suggests further refinements to your search. For example, once it’s listed all your friends who like Beyoncé, Facebook may suggest refining those results down to just the single ladies.

It doesn’t show people things they couldn’t see before; anything your privacy settings allow people to see they can now find more easily via Graph Search.

Graph Search is in “beta,” being rolled out for a limited number of English-language users for now. Eventually everyone will get it.

The Trouble with Graph Search

A feature is neither good nor bad by itself. What matters is how people actually use it.

Facebook Graph Search will usher in some new behaviors. In fact, that’s the whole point. Facebook poached Google heavyweights Lars Rasmussen and Tom Stocky to create Graph Search and spent an entire year building it, mainly to make Facebook more engaging so that people would spend a lot more time using Facebook.

So what are these new behaviors?

For starters, Facebook users and social media experts will start talking about SGO — Social Graph Optimization.

Want to get recruited and hired by a better company? You’ll need to change your profile and behave in a different way on Facebook so that when headhunters use Graph Search, your name will come up.

Want to be a more effective leader, find a new girlfriend, avoid getting fired or find new customers for your small business? You’ll need to “tweak” your Facebook activities so you’ll achieve the desired outcomes when people search for things. Experts will offer advice on using Facebook for reputation management.

Another new behavior is that an active, more technical and savvy minority will go into the Facebook privacy settings and lock things down better. Smart users will also keep Graph Search in mind when they’re liking things, commenting and generally interacting on Facebook.

The majority of users won’t change their settings, and it will be easy to violate their privacy in a variety of powerful ways.

Experimenting with Graph Search will become a hobby for many people — they’ll try different combinations of criteria just for fun. They’ll slice and dice and refine different search criteria to discover random things about people.

Unethical users will use Graph Search to stalk, con and even rob people. They’ll be able to search for certain types of people who hang out in specific locations, and also learn about people’s relationships much more easily than they can now. It won’t be hard to use Graph Search to figure out when people will be away from home.

As Gizmodo demonstrated with a series of profane and offensive test searches, it will be easy using Graph Search to bring up ill-advised past use of the Facebook “Like” button.

In fact, it’s easy to predict that this will become a widespread Internet meme. Jokesters will search for things like “People who like getting drunk and passing out,” then post the resulting screenshots of the profile photos and names. The best of these (and by best I mean worst) will circulate virally.

If Facebook is good for bullies, Facebook plus Graph Search will make Facebook great for bullies.

And More Trouble with Graph Search

The quality of any search system depends on the quality of available data in the index. And I think Facebook will have trouble with data.

Facebook already broadcasts your Likes to your family and friends, and often gets this wrong. Vegetarians are advertised on Facebook as “Liking” bacon, for example. Even people who have died are showing up as having recently “Liked” something on Facebook.

The other problem Facebook struggles with is fake Likes, where people click Like on things inaccurately, erroneously or fraudulently. The BBC did a test by setting up a fake company and got real likes from fake profiles and fake likes from real profiles.

Graph Search will compound the problem. People who want attention, or to spam or flirt or otherwise get noticed will “Like” everything in sight in the hopes of getting in on the results of as many searches as possible, thereby mucking up the index with junk data.

In general, however, Graph Search makes Facebook a more powerful version of everything Facebook already is.

That’s a good thing or a bad thing depending on what you already think of Facebook.

Subscribe to Data Insider

Learn the latest news and best practices about data science, big data analytics, artificial intelligence, data security, and more.

Similar articles

Get the Free Newsletter!

Subscribe to Data Insider for top news, trends & analysis

Latest Articles