Reducing Hubness: A Cause of Vulnerability in Recommender Systems.
Kazuo HaraIkumi SuzukiKei KobayashiKenji FukumizuPublished in: SIGIR (2015)
Keyphrases
- recommender systems
- collaborative filtering
- user profiles
- user profiling
- user modeling
- cold start problem
- information overload
- user preferences
- matrix factorization
- information filtering
- data mining
- decision making
- website
- risk assessment
- recommendation algorithms
- information retrieval
- neural network
- recommendation quality
- trust propagation