Slope One Predictors for Online Rating-Based Collaborative Filtering

By Daniel Lemire, Anna Maclachlan. SDM, 2007

727 citations recorded by the source corpus.

Rating-based collaborative filtering is the process of predicting how a user would rate a given item from other user ratings. We propose three related slope one schemes with predictors of the form f(x) = x + b, which precompute the average difference between the ratings of one item and another for users who rated both. Slope one algorithms are easy to implement, efficient to query, reasonably accurate, and they support both online queries and dynamic updates, which makes them good candidates for real-world systems. The basic slope one scheme is suggested as a new reference scheme for collaborative filtering. By factoring in items that a user liked separately from items that a user disliked, we achieve results competitive with slower memory-based schemes over the standard benchmark EachMovie and Movielens data sets while better fulfilling the desiderata of CF applications.

Subfields: Information Systems, Signal Processing

View Paper: https://www.semanticscholar.org/paper/2d38f623c7c904ecc4022443e1c22f379e89d653

How researchers see this paper

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

In Information Systems

In Signal Processing

Citations in the corpus

Explicit factor models for explainable recommendation based on phrase-level sentiment analysis (2014, methodology): “SlopeOne: A neighborhood-based CF algorithm with easy implementation and comparable or better performance than user- or item-based approaches [16].”
Comparison of collaborative filtering algorithms (2011, methodology): “The slope one algorithms [Lemire and Maclachlan 2005] are based on predictors of the form f (x) = x + b, therefore, simpler than those used in the regression-based algorithm.”
Recommendations in location-based social networks: a survey (2015, neutral): “The traditional CF models can be divided into two subgroups: 1) user-based models, such as [44], that use similarity measures between each pair of users; and 2) item-based models, such as [53], that use similarity measures between each pair of items (media content, activities, etc.”
Analysis of the Mean Absolute Error (MAE) and the Root Mean Square Error (RMSE) in Assessing Rounding Model (2018, methodology): “We compare the following algorithms with Basic Rounding approach: user-based kNN (using Pearson similarity and the neighborhood size of 30 which is suggested as the optimal value in literature [2]), Slope One [8], and Classical MF [9].”
Recommender Systems (2012, background): “It was shown that Slope One can outperform linear regression ( i.e. , estimation by f ( x ) = ax + b ) while having half the number of regressors [116,133].”
Methodologies for Cross-Domain Data Fusion: An Overview (2015, methodology): “There are two classes of memory-based CF models: user-based [45] and item-based [42] techniques.”
Classification, Ranking, and Top-K Stability of Recommendation Algorithms (2016, background): “Weighted slope one is an item-based approach that takes into account information from other users who rated the same item and from the other items rated by the same user (Lemire and Maclachlan 2005).”
A Survey of Collaborative Filtering Techniques (2009, methodology): “Lemire and Maclachlan [88] proposed slope one algorithms to make faster CF prediction than memory-based CF algorithms.”
Recommender Systems (2016, neutral): “Other regression-based models for recommender systems, such as slopeone predictors and ordinary least-squares methods, are proposed in [342, 620].”