Shallow Text Clustering Does Not Mean Weak Topics: How Topic Identification Can Leverage Bigram Features.
Julien VelcinMathieu RochePascal PonceletPublished in: DMNLP@PKDD/ECML (2016)
Keyphrases
- text clustering
- topic models
- text mining
- feature extraction
- image features
- information retrieval
- prior knowledge
- feature vectors
- document clustering
- text documents
- machine learning
- probabilistic model
- topic modeling
- question answering
- document collections
- text categorization
- co occurrence
- similarity measure
- wordnet
- latent dirichlet allocation
- user feedback
- metric learning
- text data
- text classification