Functional matrix factorizations for cold-start recommendation

By Ke Zhou, Shuang-Hong Yang, Hongyuan Zha. 2011

292 citations recorded by the source corpus.

Subfields: Computer Vision and Pattern Recognition, Information Systems, Management Science and Operations Research, Recommender Systems

How researchers see this paper

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

In Computer Vision and Pattern Recognition

In Information Systems

In Management Science and Operations Research

In Recommender Systems

Citations in the corpus

Image-Based Recommendations on Styles and Substitutes (2015, background): “Creating a labeled dataset is particularly onerous when modeling pairwise distances because the number of annotations required scales with the square of the number of elements.”
Cross-Domain Recommendation: An Embedding and Mapping Approach (2017, background): “Unfortunately, the rating matrix is often highly sparse in practical scenarios, forming a hard barrier for MF to be widely used in realistic recommender systems [Schein et al., 2002; Lin et al., 2013; Zhou et al., 2011].”
A survey of active learning in collaborative filtering recommender systems (2016, methodology): “• Functional Matrix Factorization (fMF) [90]: In this method, similarly to other decision tree methods, a decision tree is constructed for managing the interviews with the users and elicit their ratings.”
Interactive collaborative filtering (2013, methodology): “The items used in the interview can be selected based on measures such as popularity, entropy and coverage [30, 31], or a decision tree to partition the users [41, 17].”
Inferring Networks of Substitutable and Complementary Products (2015, background): “Although it is promising that we are able to recommend substitutes and complements from the text of product reviews, an obvious question remains as to what can be done for new products, that do not yet have any reviews associated with them, known as the ‘cold-start’ problem in recommender systems [1”
The FacT: Taming Latent Factor Models for Explainability with Factorization Trees (2019, methodology): “Some systems [27, 40] combine decision tree learning with matrix factorization to extract a list of interview questions for solving the cold-start problem in recommendation.”
Learning multiple-question decision trees for cold-start recommendation (2013, methodology): “Our linear model is TL(xi) = ZLx̄i, TD(xi) = ZDx̄i and TU (xi) = ZU x̄i, where ZL, ZD and ZU are (t + 1) × k matrices for each individual node in the next level.”
Facing the cold start problem in recommender systems (2014, background): “In Zhou, Yang, and Zha (2011), the authors present functional matrix factorization (fMF). fMF constructs a decision tree for the initial interview (each node being an interview question) enabling the RS to query the user adaptively.”
Incorporating Phrase-level Sentiment Analysis on Textual Reviews for Personalized Recommendation (2015, methodology): “The research of personalized recommendation can be generally classified into Content-based [20], Collaborative Filtering (CF)-based [26] and Hybrid approaches [3].”
LARA: Attribute-to-feature Adversarial Learning for New-item Recommendation (2020, background): “One of the most challenging problem in existing recommender systems is the cold-start issue [42].”
Recommender Systems (2016, unknown): “Methods for addressing cold-start issues in latent factor models are discussed in [676].”