The Algorithm

The perfect algorithm for sorting posts on social media timelines. A powerful sort that almost magically shows every user what they want to see based on their interests. Free to use and adapt to your favorite datastore!

Concept

Every social media network has a different algorithm to show its users content to eachother. From Twitter and Facebook to Tiktok and Tumblr, there's a highly polished, data driven set of rules to make sure each user sees the right posts. This project creates a sort of "grand unified algorithm" that demonstrably and uncorruptably adapts these algorithms to the absolute pinnacle of content selection.

The Algorithm Itself

Prerequisite

Case α

Suppose each User u follows n other Users f.

All f|u own >=0 Posts p.

Each p has a Timestamp z.

The Timeline t is the result of all p owned by f.

Case β

Suppose each User u owns >=0 Posts p.

Each p has a Timestamp z.

The Timeline t is the result of all p in the system.

Selection

In Reverse Precedence:

Limit the selection of p from t to Total number of p to show per loading call x.
As a, for the current time c, order p by z in descending order.
Return a as the call result.

Rendering

Display each p in returned order from a to the End User, ordered from top of the viewport to bottom.

Caveats

End Users may not initially understand this selection of Posts. Most social media users expect a semi-random display of Posts and having a predictable selection of content from Users they expect, in the order they were posted may confuse them.
End Users may see Posts from Users they assumed died or left your site entirely.
You may see a stark decrease in engagement with your Support team, specifically a lack of messages complaining about not seeing friends posts. Not to worry, this is normal.

Contributing

You can help by not changing this algorithm at all, not using any other algorithm and by listening to the feedback your users give you.

License

bredmor/the-algorithm

ErrorLooks like something went wrong!

About

Resources

License

Uh oh!

Stars

Watchers

Forks

Error
Looks like something went wrong!