Grounded Feature Selection for Biomedical Relation Extraction by the Combinative Approach.
Sung Jeon SongGo Eun HeoHa Jin KimHyo Jung JungYonghwan KimMin SongPublished in: DTMBIO@CIKM (2014)
Keyphrases
- relation extraction
- feature selection
- information extraction
- entity extraction
- manually annotated
- automatic extraction
- named entities
- semantic features
- question answering
- semantic role labeling
- domain specific
- named entity recognition
- semantic relations
- text classification
- natural language processing
- text categorization
- broad coverage
- machine learning
- text mining
- reference resolution
- feature set
- feature space
- dependency trees
- annotated corpus
- classification accuracy
- model selection
- knowledge base
- tree kernels
- knowledge acquisition
- training data
- support vector
- feature extraction
- natural language
- conditional random fields
- structured data