# benfred/implicit

Fast Python Collaborative Filtering for Implicit Feedback Datasets

Repository: https://github.com/benfred/implicit
Canonical: https://ross.abutalabs.com/products/implicit
Homepage: https://benfred.github.io/implicit/
Language: Python
License: MIT
License Family: permissive
Topics: collaborative-filtering, machine-learning, matrix-factorization, recommender-system, recommendation-system, recommendation
Last push: 2026-05-08T19:01:37+00:00

## Health v2 (maintenance only)
Score: 74/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 81, release rhythm 51, longevity 100
- inputs: {"age_days": 3790, "days_push": 117, "days_rel": 117, "gap_med": null, "n_releases_24m": 1}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 3815, forks 630 (observed 2026-08-28T04:08:20.858045+00:00)

## What it is
A fast Python library implementing collaborative filtering recommendation algorithms for implicit feedback datasets, including ALS, BPR, logistic matrix factorization, and item-item nearest neighbour models. It uses Cython/OpenMP for multi-threaded CPU training and custom CUDA kernels for GPU acceleration, with optional approximate nearest neighbour support via Annoy, NMSLIB, and Faiss.

## Use cases
- build a recommender system from implicit user feedback like clicks or plays
- train an ALS matrix factorization model on a sparse user-item matrix
- find similar items in a catalog using collaborative filtering
- recommend products or music to users based on interaction history
- speed up recommendation inference with approximate nearest neighbours
- train recommendation models on GPU with CUDA

## When to choose
- you have implicit feedback data (views, purchases, plays) rather than explicit ratings
- you need fast, multi-threaded or GPU-accelerated training of classic recommendation models
- you want a simple pip-installable Python library with a scikit-learn-like fit/recommend API

## When to avoid
- you need explicit-rating-based collaborative filtering or deep learning recommenders
- you work outside Python or need a hosted recommendation service
- your dataset is tiny and a simple heuristic would suffice

## Facets
- artifact type: library
- maturity: active
- function: machine-learning, data-science
- domain: machine-learning, data-science, analytics
- platform: python, windows
- tags: collaborative-filtering, recommender-system, matrix-factorization, implicit-feedback, als, bpr, cython, cuda, linux, macos, gpu

## Member repositories
- benfred/implicit (main) score 74

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:08:20.858045+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-29T18:26:33.210477+00:00, confidence not recorded.
  - readme: https://github.com/benfred/implicit (fetched 2026-08-28T04:08:20.858045+00:00, sha 84b4a4b94c9e)
  - homepage: https://benfred.github.io/implicit/ (fetched 2026-08-29T09:21:30.994922+00:00, sha ca80d884ab95)
- Data as of 2026-08-30T08:39:29.467469+00:00.