New research reveals why Facebook ads can miss the mark

Newswise – A new study from North Carolina State University offers insight into why Facebook’s targeted advertising can sometimes sound more like wild talk. Researchers were already aware that Facebook creates interest profiles for users based on each user’s activities, but the new study finds that this process doesn’t appear to take into account the context of those activities.

“For example, if you posted something about your dislike of green cheese, the algorithm used by Facebook to infer your interests would likely notice that you shared something about green cheese,” says Aafaq Sabir, lead author of a work article and a Ph.D. student at NC State. “But Facebook’s algorithm wouldn’t record the context of your post: what you’re doing do not like green cheese. As a result, you may start receiving targeted advertisements for green cheese. »

Facebook has been open about targeting advertising to individual users based on each user’s interests. It also clarified that it infers a user’s interests based on that person’s activities. However, it is unclear exactly how this process works.

“It’s well established that Facebook’s targeting algorithm often sends people ads for things they don’t care about,” Sabir says. “But it wasn’t clear Why people were getting the bad ads.

“The implications of deducting inaccurate interest on one of the largest social media platforms in the world are significant in two respects,” says Anupam Das, co-author of the paper and assistant professor of computer science at NC State. . “This inaccuracy has both economic ramifications – as it is relevant to the effectiveness of paid advertising – and privacy ramifications, as it raises the possibility of inaccurate data being shared about individuals across multiple platforms. .”

To learn more about how Facebook generates its user interest profiles, researchers conducted two studies.

In the first experiment, the researchers created 14 new user accounts on Facebook. The researchers monitored the demographics and behavior of each account, and tracked the list of interests generated by Facebook for each account. (Each user can see the list of interests that Facebook has compiled for them by clicking on their advertising preferences, then “Categories used to contact you”, then “Interest categories.”)

“This first experiment allowed us to see what activities were associated with Facebook by inferring interest,” says Sabir. “And the key takeaway here is that Facebook is taking an aggressive approach to interest inference.

“Even something as simple as scrolling down a page led Facebook to determine that a user had an interest in that topic. For the 14 accounts we created for this study, we found that 33.22% inferred interest was inaccurate or irrelevant.”

“We then wanted to see if these results would be valid for a larger and more diverse group of users, which was the origin of the second experiment,” says Das.

In the second experiment, the researchers recruited 146 study participants from different parts of the world. Study participants downloaded a browser extension that allowed researchers to collect data from each participant’s Facebook account about their interests. The researchers then asked participants questions about the accuracy of the interest Facebook had deducted.

“We found that 29.3% of the interests that Facebook listed for study participants were actually uninteresting,” Das explains. “This is comparable to what we have seen in our controlled experiments.

“We also found that most of the study participants didn’t even know Facebook’s Ad Preferences Manager existed. They didn’t know there was a list of interests they could look at, or that Facebook at least provides a basic explanation of why it assigned a given interest to a user.

“That’s an interesting discovery in itself,” Das says. “Because the goal of providing all of this interest information is ostensibly to be transparent with users. But given that many users don’t even know this information is available, Facebook is falling short of that goal.

The paper, “Analyzing the Impact and Accuracy of Facebook Activity on Facebook’s Ad-Interest Inference Process,” will be presented at the 25th ACM Conference on Computer-Supported Cooperative Work and Social Computing (CSCW), to be held online November 12-16. . The article was co-authored by NC State undergraduate student Evan Lafontaine.

Comments are closed.