You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I want to explore using Core ML 3 to analyze the time you spend on different posts and the posts that you reply/like/bookmark to create a custom on-device machine learning algorithm that helps sort posts based on how much you would like it.
Also, Apple has textual sentiment analysis, so I could also add a toggle to promote positive posts.
Regular channels would remain the same, but this would essentially create some "algorithmic" channels that analyze the posts in all the feeds and suggest posts you might enjoy. I picture this as matching the standard Apple's "For You" Tab in many of their apps.
To start, I'll likely just make a standard algorithmic channel with some defaults for the weight of a time read vs reply vs like vs bookmark, and a toggle to add the positive post booster. Then over time I might be able to allow users to create some more custom settings around how the algorithm is run.
The text was updated successfully, but these errors were encountered:
I want to explore using Core ML 3 to analyze the time you spend on different posts and the posts that you reply/like/bookmark to create a custom on-device machine learning algorithm that helps sort posts based on how much you would like it.
Also, Apple has textual sentiment analysis, so I could also add a toggle to promote positive posts.
Regular channels would remain the same, but this would essentially create some "algorithmic" channels that analyze the posts in all the feeds and suggest posts you might enjoy. I picture this as matching the standard Apple's "For You" Tab in many of their apps.
To start, I'll likely just make a standard algorithmic channel with some defaults for the weight of a time read vs reply vs like vs bookmark, and a toggle to add the positive post booster. Then over time I might be able to allow users to create some more custom settings around how the algorithm is run.
The text was updated successfully, but these errors were encountered: