Self-Training Pre-Trained Language Models for Zero- and Few-Shot Multi-Dialectal Arabic Sequence Labeling.
Muhammad KhalifaMuhammad Abdul-MageedKhaled ShaalanPublished in: EACL (2021)
Keyphrases
- language model
- sequence labeling
- pre trained
- conditional random fields
- language modeling
- probabilistic model
- dependency parsing
- n gram
- speech recognition
- information retrieval
- document retrieval
- training data
- retrieval model
- query expansion
- structured prediction
- named entity recognition
- training examples
- test collection
- semi supervised learning
- semi supervised
- training set
- machine learning
- generative model
- relevance model
- video sequences
- video data
- co training
- feature extraction
- prior knowledge
- computer vision