As with Google and other search engines, there is a difference between Facebook search predictions and research results. Just because “in bikini” is a suggestion when searching for “photos of girlfriends” does not mean that Facebook trains its algorithm to catalog your photos of swimsuits. (That’s not to say Facebook’s AI is unable to process the images, it is. The company claims, for example, that it automatically detects and removes around 96% of adult and nude content that violates her rules. Skinny Speedo pics don’t get caught. in this net for a reason.) When I searched for “pics of my girlfriends in bikinis,” Facebook returned a weird series of images that didn’t show up in only women, but none of them wore swimsuits.
“Facebook search predictions represent what people can search for on Facebook, but do not necessarily reflect actual content on Facebook,” a spokesperson said in a statement. “We know that just because something doesn’t violate our community standards doesn’t necessarily mean people want to see it, so we’re constantly working to improve research to make sure predictions are relevant to people. “
Predictions are generated based on a number of things, including the overall popularity of the search term. This means they’re essentially a window to what other users are looking for on Facebook, which can be illuminating. When I searched for “girlfriends”, for example, some of the predictions included “who are single”, “with benefits” and “over 50”. A search for “dating” brought up suggestions such as “for the 40s, 50s, 60s and beyond” as well as “singles near me”. Unsurprisingly, it seems older people often use Facebook search to try and find dates. (The percentage of Americans over 65 who report using Facebook has doubled since 2012, according to the Pew Research Center.)
For the same reason, Facebook’s search predictions also reflect major news events. When I searched for “Oakland teachers” on Friday, the main prediction was “strike,” as nearly all of the 2,300 educators in the city of California began protesting on Thursday.
Facebook search predictions are also uniquely personalized for each user, based on pages you’ve liked, groups you’ve joined, current city you’ve provided, connections, location data, and locations. posts from the news feed and search results you engaged with in the past. Facebook search is therefore much more personalized than Google search, and for good reason. This adaptation is why when you start typing a name like “Jeff”, Facebook predicts that you are looking to search for someone you know rather than, say, Jeff Bezos.
To look for results are also personalized based on your social connections and your activity on Facebook. Again, if you search for “Jeff,” Facebook will likely match people you’re already friends with and people you have common friends with. This feature makes it easy for users to find a person they have recently met and add them to Facebook. But that doesn’t mean that the social network’s algorithm-generated search results aren’t sometimes weird. The search for “photos of my friends” produced a handful of seemingly random photos of women I know, some of which were uploaded years ago. It’s not necessarily clear why I was shown certain images rather than others.
And in De Ceukelaire’s case, a search for “photos of my male friends” returned nothing at all. When I tried it, the search did not produce any pictures of my friends, like when searching for girlfriends. Instead, I was shown a random assortment of memes. “I’ve had what are presumably male dogs and two male-themed cartoons, including one warning men against peeing outside in the polar vortex,” Melissa Locker wrote in Fast business. A Facebook spokesperson said the experiment is “a bug that we are working to fix.”