TRIP: Triangular Document-level Pre-training for Multilingual Language Models.
Hongyuan LuHaoyang HuangShuming MaDongdong ZhangWai LamFuru WeiPublished in: CoRR (2022)
Keyphrases
- language model
- document level
- language modeling
- query expansion
- n gram
- document retrieval
- probabilistic model
- test collection
- information retrieval
- retrieval model
- cross lingual
- passage retrieval
- sentence level
- pseudo relevance feedback
- sentiment classification
- language independent
- training set
- translation model
- query terms
- relevance model
- cross language
- vector space model
- text categorization
- text classification
- digital libraries