G-MAP: General Memory-Augmented Pre-trained Language Model for Domain Tasks.
Zhongwei WanYichun YinWei ZhangJiaxin ShiLifeng ShangGuangyong ChenXin JiangQun LiuPublished in: EMNLP (2022)
Keyphrases
- language model
- language modeling
- pre trained
- n gram
- information retrieval
- retrieval model
- document retrieval
- query expansion
- speech recognition
- probabilistic model
- context sensitive
- language modelling
- ad hoc information retrieval
- test collection
- mixture model
- maximum a posteriori
- smoothing methods
- dimensionality reduction
- transfer learning
- training examples
- neural network
- relevance feedback
- high dimensional
- training data