YouTube “You” doesn’t matter much when it comes to video recommendations

The company says the dislike button doesn’t do what it thinks it does


Non-profit tech company Mozilla threw a grenade at the Google-owned YouTube social video platform in the form of a 47-page report soliciting ways of negative feedback (for example, button “Dislike”) Its users have to control viewing them as ineffective, citing thousands of user experiences analyzed. YouTube dismisses the criticism, saying that the methodology does not take into account how the controls work in reality.

ANDROIDPOLICE TODAY’S VIDEO

in his reportMozilla said it collected 22,722 people’s experiences through a browser extension it created Reporter regrets. The extension creates a visible “Stop Recommendation” button above the video thumbnails on the site. However, specifically for the corresponding search group, the signal sent by Mozilla to YouTube may differ between the various forms of feedback the site currently offers – none (for control purposes), disliked, disinterested (in the video), does not recommend the channel, Remove from history.

Of the four actual actions, the most effective method was “Do not recommend channel,” preventing 43% of unwanted recommendations. Users who removed videos from their watch history had a 29% success rate while the dislike and disinterest rate increased by 12% and 11%, respectively. For reference, the control group had a poor video recommendation rate of about 40%. Subtracting video recommendations from the same channel, this number drops to nearly 30%. The overall rate of bad recommendations across all comment modes has decreased over time by at least 10 percentage points.

In a pre-trial survey of 2,758 people, Mozilla said participants were generally skeptical about whether their actions had any effect at all on the algorithm that drives their video suggestions.

The organization concluded that YouTube should be clearer about what comment tools do to alter users’ experiences on the site and to provide more user controls that put too much weight on the algorithm.

In a statement to the edgeYouTube spokeswoman Elena Hernandez says Mozilla’s report relies on creating definitions that don’t reflect what commenting tools actually do: the “Disinterested” option removes this video only as a factor for future recommendation while the “Don’t recommend channel” option removes the entire channel.

Moreover, its algorithms are programmed regardless of entire topics or perspectives to discourage users from forming echo chambers.

Hernandez also states that the company has expanded its access to the Data API through the YouTube Research Program.

Mozilla countered this point by saying that its research appeared in metrics that the API wasn’t providing. He also supports his view that user feedback is generally less of a priority when YouTube initiates a policy change.

When a mechanism like YouTube’s algorithm-based video recommendation system works in an opaque way, Mozilla claims, people will use the mechanisms given to them to try to control it.