ERNIE-Gram: Pre-Training with Explicitly N-Gram Masked Language Modeling for Natural Language Understanding.
Dongling XiaoYu-Kun LiHan ZhangYu SunHao TianHua WuHaifeng WangPublished in: NAACL-HLT (2021)
Keyphrases
- language modeling
- n gram
- natural language understanding
- language model
- knowledge representation
- semantic analysis
- language independent
- text classification
- statistical language modeling
- natural language
- language modelling
- cross lingual
- natural language processing
- part of speech
- information retrieval
- word segmentation
- retrieval model
- query expansion
- probabilistic model
- dialogue system
- expert finding
- document retrieval
- test collection
- relevance model
- hidden markov models
- training set
- bayesian networks
- finite state transducers
- text classifiers
- data mining