lightfm

A Python implementation of LightFM, a hybrid recommendation algorithm.

View on GitHub↗Upstream lyst/lightfm↗

Builder

lyst

lyst • individual

Stars

5,072

Using upstream star count

Forks

727

Using upstream fork count

Open Issues

Activity Score

0/100

0 commits in 30d

Created

Jul 30, 2015

Project creation date

README Summary

LightFM is a Python implementation of a hybrid matrix factorization model that can incorporate both collaborative and content-based information for recommendations. It supports both implicit and explicit feedback and can handle cold-start problems by leveraging user and item features. The library is designed to be fast and memory-efficient, with implementations in both Python and Cython.

AI Dev Skills

Unmapped

Tags

Hybrid Recommendation SystemsCollaborative FilteringContent-Based FilteringMatrix FactorizationImplicit Feedback LearningFeature Engineering for RecommendersGradient Descent OptimizationSparse Matrix OperationsHybrid AI SystemsContent RecommendationCross-selling OptimizationMedia & EntertainmentE-commerceStreaming PlatformsPersonalized News FeedProduct RecommendationsMarketplace PlatformsPersonalization at ScaleTabularSelf-hostedItem-to-Item SimilarityCloud APISocial MediaOn-premiseCold Start RecommendationsContent PublishingMovie/Music RecommendationPython

Recent Activity

Updated 1 years ago

7 Days

30 Days

90 Days

Quality

production

Quality: medium
Maturity: production

PM Skills

Developer Platform

Languages

Python100.0%

Timeline

Project created: Jul 30, 2015
Forked: Mar 22, 2026
Your last push: 1 years ago
Upstream last push: 1 years ago
Tracked since: Jul 24, 2024

Similar Repos

pgvector cosine similarity · $0

Loading…