A Survey of Collaborative Filtering Techniques
By Xiaoyuan Su, Taghi M. Khoshgoftaar. 2009
4,031 citations recorded by the source corpus.
As one of the most successful approaches to building recommender systems, collaborative filtering (CF) uses the known preferences of a group of users to make recommendations or predictions of the unknown preferences for other users. In this paper, we first introduce CF tasks and their main challenges, such as data sparsity, scalability, synonymy, gray sheep, shilling attacks, privacy protection, etc., and their possible solutions. We then present three main categories of CF techniques: memory-based, modelbased, and hybrid CF algorithms (that combine CF with other recommendation techniques), with examples for representative algorithms of each category, and analysis of their predictive performance and their ability to address the challenges. From basic techniques to the state-of-the-art, we attempt to present a comprehensive survey for CF techniques, which can be served as a roadmap for research and practice in this area.
Subfields: Computer Science Applications, Information Systems, Transportation
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
In Computer Science Applications
- Hao Wang
- John Riedl
- J. Konstan
- Harald Steck
- Wu-Jun Li
- Todd Kulesza
- Michael D. Ekstrand
- Tat-seng Chua
In Information Systems
In Transportation
- Yu Zheng
- Aixin Sun
- Anne-Marie Kermarrec
- Hui Xiong
- Yongfeng Zhang
- Vincent Wenchen Zheng
- Xing Xie
- Yanjie Fu
Citations in the corpus
- Collaborative Topic Regression with Social Matrix Factorization for Recommendation Systems (2012, background): “Collaborative Filtering (CF), which automatically predicts the interests of a particular user based on the collective rating records of similar users or items, has been extensively studied in the literature (Hu et al., 2008; Salakhutdinov & Mnih, 2008; Su & Khoshgoftaar, 2009).”
- Deep Neural Networks for YouTube Recommendations (2016, background): “YouTube is the world’s largest platform for creating, sharing and discovering video content.”
- Federated Collaborative Filtering for Privacy-Preserving Personalized Recommendation System (2019, methodology): “Particularly, CF utilizes user data to build models and generate recommendations for users, in different contexts [5].”
- A Hybrid Collaborative Filtering Model with Deep Structure for Recommender Systems (2017, methodology): “Nevertheless, CF-based methods are often preferred to content-based methods because of their impressive performance (Su and Khoshgoftaar 2009).”
- Explainable recommendation with fusion of aspect information (2018, methodology): “Collaborative Filtering (CF) based approaches [4, 15, 34, 38] make rating prediction using a decomposition of a user-item rating matrix, and fall into two categories, the memory based CF [6, 11, 12] and the model based CF [16, 17, 31].”
- A Neural Influence Diffusion Model for Social Recommendation (2019, background): “By providing personalized item suggestions for each user, recommender systems have become a cornerstone of the E-commerce shopping experience [2, 34].”
- Online Learning: A Comprehensive Survey (2018, background): “Several survey paper provides detailed introduction to the regular CF techniques [180, 181].”
- Social Collaborative Viewpoint Regression with Explainable Recommendations (2017, methodology): “Unlike content-based filtering strategies [23] that predict ratings using the analysis of user profiles, collaborative filtering [39] methods, either memory-based CF or model-based CF, predict ratings using user-item ratings matrices.”
- A survey of active learning in collaborative filtering recommender systems (2016, methodology): “Several approaches have been proposed and used for recommendation generation [22,23,21,17]:”
- CityTransfer (2018, methodology): “Collaborative filtering (CF) [34] uses the known preferences of a group of users to make recommendations or predictions of the unknown preferences for other users.”
- Machine learning approaches and databases for prediction of drug–target interaction: a survey paper (2020, background): “Thesemethodstakeadvan-tageoftherecommendersystemapproaches[75,76],whileusing both chemical and genomic information is optimal for the DTI prediction problem.”
- Artificial intelligence in recommender systems (2020, methodology): “CF-based techniques are classified into two categories [18]: memory-based CF andmodel-based CF.”
- 2015 Ieee International Conference on Big Data (big Data) Task-based Recommendation on a Web-scale (neutral): “The research on Personalized Recommender Systems (PRS) [1] has long been focusing on Collaborative Filtering (CF)-based [2] and the related techniques, without explicit mining, modeling, and prediction of the underlying tasks and diverse needs that inherently dictate users' browsing behaviors.”
- Link Prediction in Complex Networks: A Survey (2010, background): “tering 2 framework [30]. 2 Collaborative filtering is the process of filtering for information or patterns using techniques involving collaboration among multiple agents, viewpoints, data sources, etc. [29] 5 Node similarity can be defined by using the essential attributes of nodes: two nodes are consi”
- Learning Image and User Features for Recommendation in Social Networks (2015, methodology): “CF [28] recommends images to users based on the images shared by other users with similar interest.”
- Reinforcement Learning based Recommender Systems: A Survey (2021, methodology): “Plenty of research has been done in the field of RSs and a plethora of survey papers have been published, including RSs [12], collaborative filtering [33, 34], hybrid methods [35], multi-media RSs [36, 37], explainable recommendation [38], and article RSs [39], to name a few.”