YouTube is continuing with the improvements to the recommendation engine in an effort to offer users the right videos along with reducing the reach of videos that spread misguide them.
Since these videos do not quite violate community guideline, YouTube instead of removing is looking another way to limit the recommendations of these borderline content as well as content that misguide in harmful ways.
For example “videos involved in click-baiting with misleading titles and descriptions, videos that promote a phony miracle cure for serious illness, claiming the earth is flat or making blatantly false claims about historical events like 9/11.”
According to YouTube, less than one percent of video content qualifies as borderline content on its platform.
Whle this change will only affect recommended videos, users, however, can still access these type of videos, and when relevant YouTube may also recommend subscribers of channels that publishes borderline content and when specifically searched for.
YouTube says, it recommends over 200 million videos on its homepage every day, that are sourced from a broader set of topics.
In addition to hundreds of improvements made to the recommendations, this change will leverage a combination of machine learning (ML) along with human evaluators to make the recommendation system more accurate.
“These evaluators are trained using public guidelines and provide critical input on the quality of a video.”
The new recommendations system change will initially affect a small set of videos in the US.
And, since it’s a gradual change, YouTube says it will roll out the change to more countries over time.