Feature selection based on word-sentence relation.
Péter SchönhofenAndrás A. BenczúrPublished in: ICMLA (2005)
Keyphrases
- feature selection
- dependency relations
- sentence level
- syntactic information
- pointwise mutual information
- text corpus
- word level
- training corpus
- noun phrases
- syntactic analysis
- text classification
- text categorization
- natural language
- mutual information
- sentence similarity
- sentiment analysis
- information gain
- co occurrence
- feature extraction
- stop words
- document level
- syntactic categories
- lexico syntactic
- word frequency
- document frequency
- classification accuracy
- machine learning
- word recognition
- n gram
- feature selection algorithms
- maximum entropy classifier
- feature set
- knn
- linguistic knowledge
- part of speech
- parallel corpus
- support vector
- word order
- sentence pairs
- named entities
- word sense disambiguation
- ambiguous words
- term frequency
- sentiment classification
- feature space
- support vector machine
- linguistic features