Feature selection via maximizing global information gain for text classification.
Changxing ShangMin LiShengzhong FengQingshan JiangJianping FanPublished in: Knowl. Based Syst. (2013)
Keyphrases
- global information
- text classification
- feature selection
- text categorization
- bag of words
- structural information
- naive bayes
- n gram
- contextual information
- knn
- text mining
- mutual information
- machine learning
- text documents
- globally consistent
- support vector machine
- support vector
- feature weighting
- global context
- global knowledge
- classification accuracy
- feature space
- feature engineering
- text classifiers
- labeled data
- term frequency
- semantic features
- dimensionality reduction
- supervised feature selection
- feature set
- feature extraction
- search engine
- k nearest neighbor
- databases
- distributional clustering
- feature reduction
- data cleaning
- feature selection algorithms
- feature subset
- prior information
- microarray
- semantic information
- data analysis
- unlabeled data