Parameter estimation based on EM algorithm for antifolksonomical item recommendation system in social bookmarking.
Akira SasakiTakamichi MiyataYoshinori SakaiYasuhiro InazumiAki KobayashiPublished in: IADIS AC (1) (2009)
Keyphrases
- social bookmarking
- parameter estimation
- em algorithm
- personalized recommendation
- expectation maximization
- maximum likelihood
- collaborative filtering
- recommender systems
- recommendation systems
- user interests
- mixture model
- maximum likelihood estimation
- web users
- user feedback
- gaussian mixture model
- generative model
- gaussian mixture
- likelihood function
- recommendation algorithms
- approximate inference
- maximum a posteriori
- user preferences
- posterior distribution
- probabilistic model
- probability density function
- user experience
- spam detection
- gibbs sampling
- maximum likelihood estimates
- feature selection
- parameter learning
- least squares
- image segmentation
- web pages