Recommender Systems Workshop

Hi all!

• Please help us by keeping this list updated :)
• Main homepage: Recommender Systems Workshop

Recommender Systems and Personalization Datasets

collection of datasets that have been collected for research

• by the group of Julian McAuley at UCSD
• Download

MovieLens (Movies)

One of the most popular datasets for collaborative filtering

• Time range: 1995–2019
• Type of data: Explicit ratings (1–5 stars)
• Size: 100K, 1M, 10M, 25M ratings
• Download

Amazon Product Reviews (E-commerce)

Large-scale dataset of product reviews and ratings

• Time range: 1996–2018
• Type of data: Explicit ratings (1–5 stars), reviews
• Size: 233M+ reviews across categories
• Download

Netflix Prize (Movies)

Classic dataset from the Netflix Prize competition

• Time range: ~1998–2005
• Type of data: Explicit ratings (1–5 stars)
• Size: 100M+ ratings from 480K users
• Download

Yelp Open Dataset (Local Businesses)

Large dataset for local business recommendation

• Time range: 2004–2022
• Type of data: Explicit ratings (1–5 stars), reviews
• Size: 8.6M reviews from 1.3M users
• Download

Spotify Million Playlist Dataset (Music)

Dataset for playlist-based music recommendation

• Time range: 2010–2017
• Type of data: Implicit interaction (playlists)
• Size: 1M playlists, 2M unique tracks
• Download

MIND (News Recommendation)

Large-scale dataset of MSN News click interactions

• Implicit feedback (clicks on news articles)
• Includes article titles, abstracts, categories
• Download

Yahoo! Music Ratings (Music)

Explicit song and album ratings from Yahoo! Music users

• Explicit ratings (scale 0-100)
• Part of KDD Cup 2011 challenge
• Download