https://github.com/abirusabil123/discover
An open source social media platform to discover great links without any implicit personalized recommendation algorithms
https://github.com/abirusabil123/discover
android discover no-personalize no-recommendations social-media website
Last synced: 5 months ago
JSON representation
An open source social media platform to discover great links without any implicit personalized recommendation algorithms
- Host: GitHub
- URL: https://github.com/abirusabil123/discover
- Owner: abirusabil123
- License: other
- Created: 2025-07-19T05:12:46.000Z (12 months ago)
- Default Branch: main
- Last Pushed: 2026-01-29T14:36:40.000Z (5 months ago)
- Last Synced: 2026-01-30T03:48:24.728Z (5 months ago)
- Topics: android, discover, no-personalize, no-recommendations, social-media, website
- Language: Kotlin
- Homepage: https://discoverall.space/
- Size: 1.15 MB
- Stars: 1
- Watchers: 0
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# 🌏 Discover - Open source no implicit personalized recommendation algorithm social media
A modern web application that helps you discover amazing links.
Features include random link loading, app usage tracking, likes/dislikes functionality, and the ability to add new links.
We host the backend locally using reproducible containers. We also have a frontend only fallback to maintain availability.
## Avoid Implicit Personalized Recommendation Algorithms
A social media with no personalized recommendation algorithms. So no more echo chamber radicalization.
There are two ways to end personalized recommendation algorithms:
1. Block user data. No user data means no user personalization.
2. No recommendation algorithm. User data is not used for implicit (without user choice like in subscriptions) personalization.
Any of the two is sufficient. My discover link does both.
Discover something new everyday. Good mood and learning ahead!
## Background
It is like tiktok for web links.
Completely self reliant and you can share links with the whole world, upvote them, downvote them, and it shows these counts including visits. To prevent abuse, bad links are automatically removed if they get many downvotes that are more than 80% of total votes.
This is my contribution so you don't use personalized recommendation algorithms social media. They cause echo chamber radicalization.
Let's go back to before 2005, before EdgeRank was released.
## Run frontend for local testing
```
npx serve . -l 44631
```
## Format code in VS Codium
```
Ctrl+Shift+I
```
## Add LICENSE header to new file
```
python addHeader.py
```