Seeing the wood for the trees: a contrastive regularization method for the low-resource Knowledge Base Question Answering.
Junping LiuShijie MeiXinrong HuXun YaoJie (Jack) YangYi GuoPublished in: NAACL-HLT (Findings) (2022)
Keyphrases
- question answering
- regularization method
- knowledge base
- image restoration
- natural language processing
- regularization parameter
- question classification
- information extraction
- decision trees
- information retrieval
- syntactic information
- named entities
- passage retrieval
- expert systems
- knowledge representation
- natural language
- tree structure
- inverse problems
- qa clef
- partial differential equations
- relation extraction
- answering questions
- question answering systems
- natural language questions
- global optimization
- cross validation
- wordnet
- answer extraction
- background knowledge
- qa systems
- pairwise
- bayesian networks
- feature selection