Paragraph-based Transformer Pre-training for Multi-Sentence Inference.
Luca Di LielloSiddhant GargLuca SoldainiAlessandro MoschittiPublished in: NAACL-HLT (2022)
Keyphrases
- structured prediction
- training set
- linguistic features
- sentence level
- training corpus
- natural language
- training samples
- power system
- bayesian networks
- fuzzy logic
- fault diagnosis
- inference process
- training examples
- semantic role labeling
- training algorithm
- neural network
- text classification
- probabilistic inference
- document level
- training phase
- part of speech
- inference engine
- semantic analysis
- training process
- belief networks
- online learning
- expert systems
- genetic algorithm