Building a ranking, recommendation, and reputation infrastructure using the EigenTrust algorithm and more
Gm Lensters! @lenster.lens @lensprotocol If I can ask you to do this 10s quick poll and help everyone understand the social graph trust in Lens, that'd be great help! Thanks!
quick poll: Which reaction types would you rate highest and lowest signal for capturing trust or engagement intent from Profile A-->B:
1 A Follows B
2 A Likes B's posts
3 A Comments on B's posts
4 A Collects B's posts
5 A Mirrors B' posts
6 A Mention's B in a post/comment
Direct comment with your answer works!
Or type in this form: tally.so/r/mBdoP1
π Introducing Choose Your Algorithm at Lenster π
Step into a new era of social media personalization with our unique 'Choose Your Algorithm' feature! πβ¨
Explore our new collection of open-source algorithms powered by @karma3labs.lens and @lenster.lens β¨
Navigate the web the way you want. ππ
Stay tuned more algorithms are on the way for ya!! π
ofc voted for Following by Karma3Labs as it's the first, and so far only, personalised feed on Lens :-)
K3L has great high-level description of the algo here: docs.karma3labs.com/developers/lens-protocol/content-recommendations cc @karma3labs.lens
We are a reputation protocol helping front-end devs to customize search and discovery of reputable contents/profiles in decentralized social and many other areas :)
We are a reputation protocol helping front-end devs to customize search and discovery of reputable contents/profiles in decentralized social and many other areas :)
We are a reputation protocol helping front-end devs to customize search and discovery of reputable contents/profiles in decentralized social and many other areas :)
Proud to be the content recommendation computation provider! Thank you @nader.lens @wagmi.lens @stani.lens
docs.lens.xyz/docs/other-apis-and-algorithms
π We're launching content discovery and recommendation APIs for @Lens Protocol - published under the Open Algorithm Standards (Lens Improvement Proposals).
π Learn more: (docs.karma3labs.com/developers/lens-protocol/content-recommendations)) π»Demo: content.lens.k3l.io/
Join the Karma3 Labs dev channel for early access: t.me/Karma3Labs
π― We're building an open ranking and recommendation layer for decentralized social media. Developers can use our open and verifiable computation infrastructure for People and Content Discovery and personalization.
π‘ Developers can choose from a variety of custom feed algorithms from Karma3 Labs' open-source repo. #DecentralizedSocialMedia
π Our aim is to help Lens developers to:
π Grow user acquisition and retention through engaging and personalized content.
π« Reduce spam users and content by using reputation signals from the social graph.
π° Reduce friction, costs and resources in managing the data and algorithm layer.
π Customizable Algorithms to Fit Your Needs π
We offer a collection of customizable 'feed' algorithms for social applications. Developers can choose the algorithms that align with their requirements and audience. Soon, they'll even be able to fine-tune parameters for specific algorithms. #CustomizableAlgorithms
π Combining Graph-based Reputation Algorithms & ML π€
Our feed algorithms leverage graph-based reputation algorithms like EigenTrust and combine them with Machine Learning techniques. This fusion allows us to create custom feeds that surface high-quality posts while reducing spam. #Reputation #MachineLearning
π Global Content and Personalized Algorithms π
Discover our current APIs providing global content and personalized algorithms:
1οΈβ£ Recent - Real-time view of the latest content across the platform.
2οΈβ£ Popular - Highlighting recent viral posts by popular profiles.
3οΈβ£ Recommended - AI-powered discovery of new and interesting posts.
4οΈβ£ Crowdsourced - Trust-based recommendations from interactions.
5οΈβ£ Following - Personalized feed based on profiles you follow. #APIs #ContentDiscovery
π Recent Strategy: Real-Time Insights π
Offers users a real-time view of the latest content shared on the platform. Posts are fetched from the Lens Public BigQuery dataset and arranged based on posting time. Stay updated with the most recent posts! #RecentPosts #RealTimeInsights
πPopular Strategy: Viral Posts by Trusted Profiles π
Highlights the most recent viral posts by trusted profiles. EigenTrust scores and publication stats are combined to determine recommendation scores for each post. Discover the latest trending content! #PopularPosts #Trending
π Recommended Strategy: AI-Powered Discovery π§
Explore new and interesting posts powered by AI and EigenTrust. The ML model classifies posts based on the features of highly engaging posts from top-scoring profiles. Unleash the power of AI in content discovery! #MLRecommendations
π‘Crowdsourced Strategy: Truly Recursive Trust-Based π
View posts that are weighted by the reputation of the interacting parties. This approach, based on Hubs and Authorities algorithm, assigns trust scores to profiles and publications. By leveraging credit trust, both publications and profiles contribute to each other's trustworthiness.
π€© Following Strategy: Personalized and Tailored to your preferences and interests. π
Tailors to user preferences and interests - this personalized algorithm surfaces content posted by the profiles they follow. With engagement as the key criterion, we collect and score posts based on publication stats and EigenTrust scores of the profile. Connect with what matters most to you!
π Useful Links:
π Ask us Anything: t.me/Karma3Labs
π Tutorials and More on Strategies Explained: docs.karma3labs.com/developers/lens-protocol/content-recommendations
β³ Demo: content.lens.k3l.io/
π€ Open API and Github: openapi.lens.k3l.io/#/
Hey Jacquest, join our telegram dev chat, let's get in touch! πt.me/Karma3Labs