InfoXLM: An Information-Theoretic Framework for Cross-Lingual Language Model Pre-Training.
Zewen ChiLi DongFuru WeiNan YangSaksham SinghalWenhui WangXia SongXian-Ling MaoHeyan HuangMing ZhouPublished in: NAACL-HLT (2021)
Keyphrases
- cross lingual
- language modeling
- language model
- translation model
- pseudo feedback
- n gram
- information retrieval
- document retrieval
- cross lingual information retrieval
- language independent
- retrieval model
- cross language
- query expansion
- probabilistic model
- machine translation
- cross language retrieval
- speech recognition
- statistical machine translation
- vector space model
- parallel corpora
- text retrieval
- context sensitive
- test collection
- bayesian networks
- retrieval effectiveness
- query terms
- pseudo relevance feedback
- transfer learning
- generative model
- query specific
- text classification
- information extraction