Seventh IEEE International Conference on Data Mining Scalable Collaborative Filtering with Jointly Derived Neighborhood Interpolation Weights

Citations in the corpus

• From amateurs to connoisseurs: modeling the evolution of user expertise through online reviews (2013, background): “As discussed in [20], few previous works had dealt with temporal dynamics, other than a few notable exceptions [2, 10, 35].”
• Social collaborative filtering for cold-start recommendations (2014, background): “While traditional collaborative filtering methods [1, 6] often provide strong recommendation performance once a user has made a few item purchases or preference judgments, they cannot be used in the cold-start setting when there are no user interactions with items.”
• Advances in Collaborative Filtering (2011, methodology): “To put this in some perspective, this result is even better than those repo rted by using hybrid approaches such as applying a neighborhood approach on resi duals of other algorithms [2, 23, 31].”
• Collaborative filtering with temporal dynamics (2010, methodology): “To put this in some perspective, this result is even better than those reported [ 1 , 10, 21] by using hybrid approaches such as applying a neighborhood approach on residuals of other algorithms.”
• Transfer Learning in Collaborative Filtering for Sparsity Reduction (2010, unknown): “Observed Without Transfer With Transfer (sparsity) AF LFM CMF CST MAE 10 (0.2”
• The BellKor solution to the Netflix Prize (2007, methodology): “The core components of the solution are published in our ICDM'2007 paper [1] (or, KDD-Cup’2007 paper [2]), and also in the earlier KDD'2007 paper [3].”
• Transfer learning in heterogeneous collaborative filtering domains (2013, methodology): “Probabilistic matrix factorization The PMF [47] or latent factorization model (LFM) [4] seeks an appropriate low-rank approximation, R = UVT , for which any missing value can be predicted by r̂ui = Uu·V Ti· , where U ∈ Rn×d , V ∈ Rm×d are user-specific and item-specific latent feature matrices, resp”
• Statistical ranking and combinatorial Hodge theory (2008, background): “For example, in the Netflix context, viewers will rate a movie on a scale of 5 stars [5]; in financial markets, analysts will rate a stock or a security by 5 classes of recommendations [4].”
• Testing for Associations between Loci and Environmental Gradients Using Latent Factor Mixed Models (2012, methodology): “One of these methods, the alternate least square method uses deterministic steps that are similar to our stochastic Gibbs sampler [32].”
• SMF: Drift-Aware Matrix Factorization with Seasonal Patterns (2019, methodology): “Other work incorporated bias terms [13], and alternating least squares [2].”
• On the stability of recommendation algorithms (2010, methodology): “5 Impact of Data Normalization In all previous experiments, before applying collaborative filtering approaches or matrix factorization method, the rating data was normalized by removing “global effects” (including the overall mean, user average, and item average) in the way that is often used in rec”
• Improving the quality of predictions using textual information in online user reviews (2013, methodology): “Latent factor models have been successfully used in several previous studies [3,2,32,35,22].”
• A Comprehensive Survey of Neighborhood-based Recommendation Methods (2011, methodology): “Following [1, 5, 10, 18], collaborative filtering methods can be grouped in the two general classes of neighborhood and model-based methods.”
• Collaborative Filtering for Implicit Feedback Datasets (2008, background): “Better scalability and improved accuracy make the itemoriented approach more favorable in many cases [2, 19, 20].”