Localized user-driven topic discovery via boosted ensemble of nonnegative matrix factorization.
Sangho SuhSungbok ShinJoonseok LeeChandan K. ReddyJaegul ChooPublished in: Knowl. Inf. Syst. (2018)
Keyphrases
- nonnegative matrix factorization
- user driven
- topic discovery
- text analysis
- topic models
- end users
- negative matrix factorization
- matrix factorization
- text classification
- data representation
- least squares
- latent dirichlet allocation
- spectral clustering
- principal component analysis
- natural language processing
- mashup
- feature selection
- information retrieval
- text mining
- expert systems
- objective function
- feature extraction
- original data
- wordnet
- nearest neighbor
- machine learning