Industry Report: Amazon.com Recommendations: Item-to-Item Collaborative Filtering

By Greg Linden, Brent Smith, J. York. 2003

3,575 citations recorded by the source corpus.

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Citations in the corpus

Learning Visual Clothing Style with Heterogeneous Dyadic Co-Occurrences (2015, background): “While this is a general concept, for our experiment, we define co-occurrence between items to be co-purchases.”
Session-based Recommendations with Recurrent Neural Networks (2015, positive): “Despite of its simplicity it is usually a strong baseline (Linden et al., 2003; Davidson et al., 2010).”
Biased assimilation, homophily, and the dynamics of polarization (2012, methodology): “We analyze three simple recommender algorithms—SimpleSALSA (Algorithm 1), SimpleICF (Algorithm 2), and SimplePPR (Algorithm 3)—that are similar in spirit to three well-known recommender algorithms from the literature: SALSA (36), Personalized PageRank (29), and itembased collaborative filtering (37)”
Recommender Systems: Introduction and Challenges (2015, unknown): “store for each customer [32].”
Where to Go Next for Recommender Systems? ID- vs. Modality-based Recommender Models Revisited (2023, background): “In the existing recommendation literature, there are countless models built entirely on user/item ID, from early item-to-item collaborative filtering [35],”
CLiMF: learning to maximize reciprocal rank with collaborative less-is-more filtering (2012, methodology): “One of the most typical CF methods for those scenarios is item-based CF [9, 15], in which an item-item similarity matrix is first computed, and users are recommended items that are most similar to their past relevant items.”
Exploring demographic information in social media for product recommendation (2015, background): “As historical transaction records on Jingdong is not available, we cannot compare our system with existing recommendation approaches relying on purchase history [1,5].”
Combining content-based and collaborative recommendations: A hybrid approach based on Bayesian networks (2010, unknown): “Contents lists available at ScienceDirect International Journal of Approximate Reasoning journal homepage: www.elsevier .com/locate / i jar Although both notions are closely related, this paper deals with the first type, i.e. rating prediction.”
Temporal recommendation on graphs via long- and short-term preference fusion (2010, methodology): “Two most famous algorithms of them are user-based collaborative filtering [18] (UserKNN) and item-based collaborative filtering [14] (ItemKNN).”
Session-Based Social Recommendation via Dynamic Graph Attention Networks (2019, methodology): “• ItemKNN [22] (A): inspired by the classic KNN model, it looks for items that are similar to items liked by a user in the past.”
Auralist: introducing serendipity into music recommendation (2012, background): “All artists can be sorted (in descending order) by the sumtotal of their similarity with items in a user’s existing history [13].”