Improvement of a Naive Bayes Sentiment Classifier Using MRS-Based Features.
Jared KramerClara GordonPublished in: *SEM@COLING (2014)
Keyphrases
- naive bayes
- naive bayes classifier
- naive bayesian classifier
- classification accuracy
- bayesian classifier
- decision trees
- bayes classifier
- classification algorithm
- feature set
- uci data sets
- text classifiers
- feature space
- text classification
- training data
- probabilistic classifiers
- logistic regression
- information gain
- bayesian networks
- text categorization
- combining classifiers
- probability estimation
- feature selection
- class probabilities
- augmented naive bayes
- probability estimates
- bayesian classifiers
- classification models
- cost sensitive
- data sets
- test instances
- bayesian network classifiers
- svm classifier
- conditional independence assumption
- highly skewed
- feature vectors
- machine learning
- feature subset
- multiple classifiers
- naive bayes models
- tree augmented naive bayes
- support vector machine
- base classifiers
- neural network
- classification error
- classifier combination
- class labels
- averaged one dependence estimators