Skip to content

Open benchmarking for collaborative filtering and representation-based recommendation

Notifications You must be signed in to change notification settings

ACnoWA/Open-CF-Benchmark

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

77 Commits
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Open-CF-Benchmark

Towards Open Benchmarking for Collaborative Filtering

ModelsDatasetsBenchmarking results

Models

For our benchmarking, we have evaluated the following models with open-source code and reproducible steps.

Publication Model Support MIPS? Paper Title
WWW'2001 ItemKNN NO Item-Based Collaborative Filtering Recommendation Algorithms
UAI'2009 MF-BPR YES BPR: Bayesian Personalized Ranking from Implicit Feedback
ICDM'2011 SLIM NO SLIM: Sparse Linear Methods for Top-N Recommender Systems
RecSys'2016 YoutubeDNN YES Deep Neural Networks for YouTube Recommendations
WWW'2017 NeuMF NO Neural Collaborative Filtering
WWW'2017 CML YES Collaborative Metric Learning
SIGIR'2019 NGCF YES Neural Graph Collaborative Filtering
WWW'2019 EASE^R NO Embarrassingly Shallow Autoencoders for Sparse Data
AAAI'2020 LR-GCCF YES Revisiting Graph based Collaborative Filtering: A Linear Residual Graph Convolutional Network Approach
SIGIR'2020 LightGCN YES LightGCN: Simplifying and Powering Graph Convolution Network for Recommendation
TOIS'2020 ENMF YES Efficient Neural Matrix Factorization without Sampling for Recommendation
2021 SimpleX YES

Datasets

For our benchmarking, we employ the following open datasets that are commonly used in previous work.

Dataset Dataset_ID Contain features? Description
AmazonBooks amazonbooks_x0 NO The preprocessed data is provided in LightGCN.
amazonbooks_x1 YES The preprocessed data is provided in ComiRec.
Yelp18 yelp18_x0 NO The preprocessed data is provided in LightGCN.
Gowalla gowalla_x0 NO The preprocessed data is provided in LightGCN.
Movielens1M movielens1m_x0 NO The preprocessed data is provided in LCFN.
movielens1m_x1 NO The preprocessed data is provided in NCF.
Movielens10M TODO
AmazonCDs amazoncds_x0 NO The preprocessed data is provided in BGCF.
AmazonMovies amazonmovies_x0 NO The preprocessed data is provided in BGCF.
AmazonBeauty amazonbeauty_x0 NO The preprocessed data is provided in BGCF.
AmazonElectronics amazonelectronics_x0 NO The preprocessed data is provided in NBPO.
CiteULike-A citeulikea_x0 NO The preprocessed data is provided in DHCF.
Taobao taobao_x0 YES The preprocessed data is provided in ComiRec.

About

Open benchmarking for collaborative filtering and representation-based recommendation

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages