A Mobile Music Recommender System based on a Two-level Genre-Rating SVM Classifier enhanced by Collaborative Filtering.
Paraskevi S. LampropoulouAristomenis S. LampropoulosGeorge A. TsihrintzisPublished in: KES IIMSS (2009)
Keyphrases
- svm classifier
- collaborative filtering
- recommender systems
- music recommendation
- support vector machine
- support vector machine svm
- genre classification
- matrix factorization
- image classification
- support vector
- user item rating
- automatic music genre classification
- recommendation quality
- user preferences
- data sparsity
- feature vectors
- information filtering
- user profiles
- mobile devices
- user ratings
- recommendation systems
- netflix prize
- support vectors
- collaborative filtering algorithms
- collaborative filtering recommender systems
- music information retrieval
- personalized recommendation
- semi supervised
- cold start
- training set
- information overload
- svm classification
- kernel function
- cold start problem
- decision function
- rating prediction
- online dating
- user interests
- product recommendation
- multi class
- informative features
- pearson correlation coefficient
- transfer learning
- rating matrix
- item based collaborative filtering
- selected features
- active user
- support vector machine classifier
- content based filtering
- implicit feedback
- recommendation algorithms
- machine learning
- ls svm
- classification accuracy
- active learning
- feature extraction
- feature selection