MASKER: Masked Keyword Regularization for Reliable Text Classification.
Seung Jun MoonSangwoo MoKimin LeeJaeho LeeJinwoo ShinPublished in: CoRR (2020)
Keyphrases
- text classification
- bag of words
- text categorization
- naive bayes
- feature selection
- keywords
- text data
- document classification
- labeled data
- n gram
- multi label
- text documents
- sentiment analysis
- text classifiers
- semantic features
- keyword search
- knn
- text mining
- unlabeled data
- prior information
- relational databases
- keyword extraction
- level set
- regularization parameter
- semi supervised learning
- training documents