Do Google+’s machine learning algorithms use mute post information

google-plusmute

I've been using Google+ frequently and have taken to muting posts in which I am disinterested. I know Google+ uses machine learning algorithms to determine what content should be displayed to me. The problem is it takes time to select to mute a post and I haven't culled my circles enough to not get a fair amount of random content. There's no shortcut for muting posts so it's much easier to just scroll past the content. But if muting a post is information Google is using to determine what posts to feed to me then I'm all for training it. On the other hand, if Google is using other features, such as amount of time a post is active on my screen, then I'm hurting myself by taking the time to mute.

Basically, I know I'll get burnt out on G+ if I keep seeing junk, so I'd like to do it right so as not to abandon social networks for years again.

Best Answer

I only see posts from people that I follow. There's a What's Hot on Google+ feature, but you can turn that off, see How do I completely disable the "Hot on Google+" feature?. I would suggest doing that, or perhaps following fewer people. Whenever one of my posts gets muted, Google+ tells me, but doesn't tell me who did the flagging. It is sort of disconcerting, so I try to avoid doing too much muting.

I don't think Google+ needs to use machine learning in Google+. I wouldn't consider "suggested people to follow" as machine learning, as it only suggests people that my existing friends (people I have in circles) follow. Of course, it is possible that Google+ is so subtle and powerful in its machine learning algorithms that it has inferred my preferences and adapted accordingly! If so, I don't think we are likely to find the details.

There is a question about ML use by Google Predictions on StackOverflow and SE Cross-validated, neither of which really could be answered with certainty. I think I answered the one on SO, which is mostly just a list of what Google Predictions does NOT use, as that is all that anyone (who doesn't work at Google) knows or is inclined to say!