HORNET: Enriching Pre-trained Language Representations with Heterogeneous Knowledge Sources.
Taolin ZhangZerui CaiChengyu WangPeng LiYang LiMinghui QiuChengguang TangXiaofeng HeJun HuangPublished in: CIKM (2021)
Keyphrases
- knowledge sources
- pre trained
- multiple knowledge sources
- domain specific
- domain knowledge
- blackboard architecture
- cultural algorithms
- semantic interpretation
- training data
- knowledge based systems
- knowledge acquisition
- semantic knowledge
- semantic constraints
- knowledge base
- natural language
- information sources
- training examples
- expert systems
- neural network
- general purpose
- learning environment
- supervised learning
- small number
- appearance variations
- state space
- data sources
- machine learning