Critically Examining the Claimed Value of Convolutions over User-Item Embedding Maps for Recommender Systems.
Maurizio Ferrari DacremaFederico ParroniPaolo CremonesiDietmar JannachPublished in: CIKM (2020)
Keyphrases
- recommender systems
- collaborative filtering
- vector space
- user profiles
- implicit feedback
- matrix factorization
- trust aware
- recommendation quality
- cold start problem
- recommendation systems
- digital images
- geodesic distance
- user modeling
- user model
- nonlinear dimensionality reduction
- graph embedding
- fourier transform
- information overload
- personalized recommendation
- user ratings
- neural network