RocketQA: An Optimized Training Approach to Dense Passage Retrieval for Open-Domain Question Answering.
Yingqi QuYuchen DingJing LiuKai LiuRuiyang RenWayne Xin ZhaoDaxiang DongHua WuHaifeng WangPublished in: NAACL-HLT (2021)
Keyphrases
- open domain question answering
- question answering
- passage retrieval
- document retrieval
- natural language
- answer extraction
- information retrieval
- training set
- audio visual
- information extraction
- natural language processing
- query expansion
- relevance model
- semantic roles
- question answering systems
- syntactic information
- multimedia