Leveraging Three Types of Embeddings from Masked Language Models in Idiom Token Classification.
Ryosuke TakahashiRyohei SasanoKoichi TakedaPublished in: *SEM@NAACL-HLT (2022)
Keyphrases
- language model
- language modeling
- language modelling
- document retrieval
- speech recognition
- n gram
- probabilistic model
- statistical language models
- information retrieval
- classification accuracy
- pattern recognition
- decision trees
- feature vectors
- retrieval model
- context sensitive
- language model for information retrieval
- feature selection
- smoothing methods
- feature extraction
- image classification
- test collection
- bayesian networks
- query expansion
- naive bayes classifier
- machine learning
- language models for information retrieval
- spoken term detection
- vector space
- translation model
- query specific
- training set
- ad hoc information retrieval
- high dimensional