RelDiff: Enriching Knowledge Graph Relation Representations for Sensitivity Classification.
Hitarth NarvalaGraham McDonaldIadh OunisPublished in: EMNLP (Findings) (2021)
Keyphrases
- classification accuracy
- feature selection
- pattern recognition
- machine learning
- decision trees
- knowledge representation
- higher level
- pattern classification
- machine learning methods
- text classification
- support vector machine
- feature extraction
- domain knowledge
- support vector
- supervised learning
- classification models
- knowledge based systems
- sensitivity analysis
- classification algorithm
- training set
- external representations
- graph theoretic
- automatic classification
- graph structure
- knowledge base
- connected components
- knowledge sharing
- preprocessing
- feature set
- image classification
- random walk
- knowledge acquisition
- directed graph
- knowledge management
- directed acyclic graph
- weighted graph
- training data
- benchmark datasets
- neural network
- classification rules
- training samples
- support vector machine svm
- shortest path
- decision rules
- domain experts