A new term-weighting scheme for text classification using the odds of positive and negative class probabilities.
Youngjoong KoPublished in: J. Assoc. Inf. Sci. Technol. (2015)
Keyphrases
- positive and negative
- class probabilities
- text classification
- naive bayes
- term frequency
- multi class
- text categorization
- probability estimates
- bag of words
- machine learning
- feature selection
- decision rules
- text mining
- tf idf
- class distribution
- text documents
- labeled data
- cost sensitive
- sentiment analysis
- class labels
- knn
- multi label
- k nearest neighbor
- classification algorithm
- test instances
- unlabeled data
- vector space model
- test collection
- pairwise
- data sets