Skip to content
Last Dictionary
Plane
Machine LearningAI Engineering & MLOps

collaborative filtering

nounAlso: CF, user-based collaborative filtering, item-based collaborative filtering, item-to-item collaborative filtering

In plain wordsRecommending things based on what people with similar tastes liked, without needing to know anything about the things themselves, as in 'people who bought this also bought…'.

  1. A family of recommender systemrecommender system, n.: Software that guesses which songs, products, posts or jobs you are most likely to want and shows those first, as when a video app lines up the next clip or a shop suggests an add-on. techniques that predict a user's preferences from the recorded preferences of many users, using only the user–item interaction matrix (ratings, purchases, clicks) rather than any description of the items. Memory-based (neighborhood) methods compute similarities directly: user-based filtering predicts from the ratings of users with similar histories, while item-based filtering, which scales better and underlies 'customers who bought this also bought', recommends items similar to those a user already liked. Model-based methods learn latent representations, most famously by matrix factorizationmatrix factorization, n.: Breaking a big table of numbers, such as which users liked which films, into two smaller tables of hidden traits, so missing entries can be filled in by multiplying them back together. and later with autoencodersautoencoder, n.: A neural network trained to squeeze its input into a compact code and then rebuild the original from that code. Learning to compress this way forces it to capture what matters in the data., graph neural networksgraph neural network, n.: A neural network for data shaped like a web of connections, such as molecules, social networks or road maps, in which each point repeatedly gathers information from its neighbors. and sequence models. The interaction matrix is extremely sparse, often with well under 1% of entries observed; the approach cannot score users or items without history (the cold start problemcold start problem, n.: The trouble a recommendation system has with brand-new users or items: with no history to learn from, it has little basis for good suggestions.) and tends to amplify already-popular items. Variants for implicit feedbackimplicit feedback, n.: Clues about what people like that come from what they do, such as clicking, buying, watching or skipping, rather than from ratings or reviews they deliberately give. treat unobserved interactions as weak negatives rather than missing ratings.

    Item-based collaborative filtering surfaced the obscure sequel because readers of the first book kept buying it.

Origin

Coined by David Goldberg, David Nichols, Brian Oki and Douglas Terry for Tapestry, a Xerox PARC system in which users annotated email and news messages for one another (Communications of the ACM, 1992). GroupLens (1994) automated rating-based prediction for Usenet news, and Amazon's item-to-item method, described by Greg Linden and colleagues in 2003, took it to e-commerce scale.

Usage note

In Tapestry, 'collaborative filtering' meant people explicitly filtering for one another; the term later shifted to automatic prediction from aggregate behavior.

Related

See also
  • recommender systemrecommender system, n.: Software that guesses which songs, products, posts or jobs you are most likely to want and shows those first, as when a video app lines up the next clip or a shop suggests an add-on.
  • matrix factorizationmatrix factorization, n.: Breaking a big table of numbers, such as which users liked which films, into two smaller tables of hidden traits, so missing entries can be filled in by multiplying them back together.
  • content-based filteringcontent-based filtering, n.: Recommending items that resemble what you already liked, judged by the items' own features, such as a film's genre, cast and plot, rather than by other people's choices.
  • cold start problemcold start problem, n.: The trouble a recommendation system has with brand-new users or items: with no history to learn from, it has little basis for good suggestions.
  • implicit feedbackimplicit feedback, n.: Clues about what people like that come from what they do, such as clicking, buying, watching or skipping, rather than from ratings or reviews they deliberately give.
  • k-nearest neighborsk-nearest neighbors, n.: Predicting something about a new example by finding the few most similar examples already known and going with their majority answer, or their average.
  • Netflix PrizeNetflix Prize, prop. n.: A 2006–2009 contest in which Netflix offered a million dollars to anyone who could predict customers' movie ratings 10 percent better than its own system. It popularized machine-learning competitions and model blending.
Contrast with
  • content-based filteringcontent-based filtering, n.: Recommending items that resemble what you already liked, judged by the items' own features, such as a film's genre, cast and plot, rather than by other people's choices.

Sources

  1. Goldberg, D., Nichols, D., Oki, B. M. & Terry, D. (1992). Using Collaborative Filtering to Weave an Information Tapestry. Communications of the ACM 35(12).
  2. Resnick, P., Iacovou, N., Suchak, M., Bergstrom, P. & Riedl, J. (1994). GroupLens: An Open Architecture for Collaborative Filtering of Netnews. Proc. ACM CSCW 1994.
  3. Linden, G., Smith, B. & York, J. (2003). Amazon.com Recommendations: Item-to-Item Collaborative Filtering. IEEE Internet Computing 7(1).

Mentioned in

recommender systemrecommender system, n.: Software that guesses which songs, products, posts or jobs you are most likely to want and shows those first, as when a video app lines up the next clip or a shop suggests an add-on., association rule learningassociation rule learning, n.: Finding 'if this, then often that' patterns in large records of transactions, such as noticing that shoppers who buy pasta and tomatoes often buy parmesan too., cold start problemcold start problem, n.: The trouble a recommendation system has with brand-new users or items: with no history to learn from, it has little basis for good suggestions., content-based filteringcontent-based filtering, n.: Recommending items that resemble what you already liked, judged by the items' own features, such as a film's genre, cast and plot, rather than by other people's choices., matrix factorizationmatrix factorization, n.: Breaking a big table of numbers, such as which users liked which films, into two smaller tables of hidden traits, so missing entries can be filled in by multiplying them back together., implicit feedbackimplicit feedback, n.: Clues about what people like that come from what they do, such as clicking, buying, watching or skipping, rather than from ratings or reviews they deliberately give., Netflix PrizeNetflix Prize, prop. n.: A 2006–2009 contest in which Netflix offered a million dollars to anyone who could predict customers' movie ratings 10 percent better than its own system. It popularized machine-learning competitions and model blending.