SIRIUS-LTG-UiO at SemEval-2018 Task 7: Convolutional Neural Networks with Shortest Dependency Paths for Semantic Relation Extraction and Classification in Scientific Papers.
Farhad NooralahzadehLilja ØvrelidJan Tore LønningPublished in: CoRR (2018)
Keyphrases
- relation extraction
- dependency trees
- scientific papers
- semantic features
- convolutional neural networks
- tree kernels
- domain specific
- automatic extraction
- information extraction
- machine learning
- text classification
- support vector
- classification accuracy
- natural language
- supervised learning
- support vector machine
- broad coverage
- artificial intelligence
- named entity recognition
- named entities
- unsupervised learning
- question answering
- scientific literature
- higher level
- feature vectors
- feature space
- feature extraction
- search engine