LST: Lexicon-Guided Self-Training for Few-Shot Text Classification.
Hazel KimJaeman SonYo-Sub HanPublished in: CoRR (2022)
Keyphrases
- text classification
- co training
- labeled data
- bag of words
- text categorization
- semi supervised learning
- naive bayes
- unlabeled data
- text mining
- machine learning
- domain specific
- feature selection
- sentiment analysis
- training set
- n gram
- video sequences
- multi label
- visual features
- natural language
- text documents
- text classifiers
- text data
- data cleaning
- video shots
- shot boundary detection
- cost sensitive
- video content
- key frames
- detection algorithm
- covariance matrices
- knn
- named entities
- sentiment classification
- multi class
- classification accuracy