Comparison of MRF and CRF for Text/Non-text Classification in Japanese Ink Documents.
Soichiro InataniTruyen Van PhanMasaki NakagawaPublished in: ICFHR (2014)
Keyphrases
- text classification
- text documents
- text data
- text classifiers
- document categorization
- markov random field
- text mining
- conditional random fields
- document classification
- text categorization
- training corpus
- labeled documents
- text representation
- text analysis
- bag of words
- automatic text classification
- text collections
- free text
- textual information
- document representation
- training documents
- information extraction
- feature selection
- keywords
- term frequency
- pairwise
- random fields
- machine learning
- textual data
- information retrieval
- digital documents
- knn
- document clustering
- sentiment analysis
- graph cuts
- image segmentation
- higher order
- text retrieval
- document collections
- latent semantic analysis
- search engine
- classify documents
- graphical models
- plagiarism detection
- document analysis
- web documents
- semantic features
- energy function
- sentiment classification
- natural language text
- textual content
- generative model
- metadata