Discussion
Loading...

Post

Log in
  • About
  • Code of conduct
  • Privacy
  • About Bonfire
dansup
dansup
@dansup@mastodon.social  路  activity timestamp 4 months ago

The Loops For You algorithm isn't some black box, not only is it open source, and will be documented, but people will be able to clear recommendation data at any time, giving you more control 馃槈

Learn more: https://joinloops.org/our-mission

How the Loops For You Algorithm works, it visualizes how the platform selects videos for users when using the For You feed. 

It shows a flowchart starting from a user request, branching based on whether the user is new (receiving a "Popular & Discovery Focus") or returning (receiving a "Personalized Focus"). Three content sources鈥擯ersonalized, Popular, and Discovery streams鈥攁re filtered for hidden creators or already-seen content and fed into a central Scoring & Ranking Engine chip icon. Below this engine, a pie chart details the scoring weights: Engagement (35%), Personalization (30%), Freshness (25%), and Creator Quality (10%). The engine outputs the final "Ranked 'For You' Feed." A feedback loop shows user actions updating interest data for future recommendations. A prominent section in the bottom right emphasizes User Control with a trash can icon, stating users can "Clear Recommendation Data at Any Time (Resets Interests & Hidden Creators).
How the Loops For You Algorithm works, it visualizes how the platform selects videos for users when using the For You feed. It shows a flowchart starting from a user request, branching based on whether the user is new (receiving a "Popular & Discovery Focus") or returning (receiving a "Personalized Focus"). Three content sources鈥擯ersonalized, Popular, and Discovery streams鈥攁re filtered for hidden creators or already-seen content and fed into a central Scoring & Ranking Engine chip icon. Below this engine, a pie chart details the scoring weights: Engagement (35%), Personalization (30%), Freshness (25%), and Creator Quality (10%). The engine outputs the final "Ranked 'For You' Feed." A feedback loop shows user actions updating interest data for future recommendations. A prominent section in the bottom right emphasizes User Control with a trash can icon, stating users can "Clear Recommendation Data at Any Time (Resets Interests & Hidden Creators).
How the Loops For You Algorithm works, it visualizes how the platform selects videos for users when using the For You feed. It shows a flowchart starting from a user request, branching based on whether the user is new (receiving a "Popular & Discovery Focus") or returning (receiving a "Personalized Focus"). Three content sources鈥擯ersonalized, Popular, and Discovery streams鈥攁re filtered for hidden creators or already-seen content and fed into a central Scoring & Ranking Engine chip icon. Below this engine, a pie chart details the scoring weights: Engagement (35%), Personalization (30%), Freshness (25%), and Creator Quality (10%). The engine outputs the final "Ranked 'For You' Feed." A feedback loop shows user actions updating interest data for future recommendations. A prominent section in the bottom right emphasizes User Control with a trash can icon, stating users can "Clear Recommendation Data at Any Time (Resets Interests & Hidden Creators).
  • Copy link
  • Flag this post
  • Block
Federation Bot
Federation Bot
@Federation_Bot replied  路  activity timestamp 4 months ago

@dansup the mobile registration process could use improvement. I just tried to register but when I went to check the mail for the registration code, the page refreshed. Now I can't register my email :/

  • Copy link
  • Flag this comment
  • Block
Federation Bot
Federation Bot
@Federation_Bot replied  路  activity timestamp 4 months ago

@dansup Very interesting! What's meant by that arrow from the engine going back to "updates interest & feedback data"?

  • Copy link
  • Flag this comment
  • Block

A small Bonfire corner on the internet

This is a small personal instance of Bonfire in the Fediverse.

A small Bonfire corner on the internet: About 路 Code of conduct 路 Privacy 路
Bonfire social 路 1.0.1 no JS en
Automatic federation enabled
Log in
  • Explore
  • About
  • Code of Conduct