Anybody in the farcaster community thinking about ML recommendation systems for a decentralized social network? Lots of interesting challenges across data collection, training, model deployment/inference, asset management, etc
I had an idea a few years back to do this sort of thing on the frontend with tensorflowjs. It was my vision for bbly.io, a firehose of news and comments (similar to Reddit) delivered over IPFS pubsub, curated clientside by models trained by users as they vote/label content. Pure static HTML social news aggregation 👍
One basic approach I'm working on with commercial cms is categorizing profiles which could either use categories for unique feeds from profiles, or categorize posts. Mostly a simple approach based on bio atm. Considering some others with recent cast and top casts
can probably just start with simple rules based systems before moving to ML
I am writing my thoughts about how we can do this in a decentralised way. RecSys are not trivial to get right and it is not an ML problem per say.. just hard to measure… Built my fair share of them in adtech, streaming, and retail and each time it is product/client/user base specific