Best of Recommendation Systems — August 2024

1. 1
   Article

   daily.dev World·2y

   Project Sauron is Live! 👁️

   Project Sauron is a new feed algorithm utilizing a cutting-edge Two Tower retrieval model, similar to those used by YouTube and Instagram, to deliver personalized content rapidly. By creating a shared embedding space for posts and users, the algorithm calculates similarities to recommend the most relevant content. Users have reported an 11% increase in reads per user and a spike in new bookmarks during the testing phase. User feedback through upvotes and downvotes helps refine these recommendations.

   921

   45
2. 2
   Article

   Towards AI·2y

   Similarity-based Based Recommendation Systems Algorithms

   Similarity-based recommendation systems use either user-user or item-item similarities to provide recommendations. User-user similarity systems compare users' preferences, while item-item systems compare items. Amazon popularized item-item similarity due to its stability over time, making it more effective in environments where user preferences change frequently. Overall, item-item approaches are preferred when there are more users than items.

   26

   1
3. 3
   Article

   KDnuggets·2y

   Building a Recommendation System with Hugging Face Transformers

   Learn to build a recommendation system using Hugging Face Transformers. This guide walks through the essential steps, from setting up the environment and processing the dataset to using embeddings and cosine similarity for accurate recommendations. It also highlights using the sentence-transformers package for transforming text into numerical vectors.

   25