Combining n-grams and deep convolutional features for language variety classification.
Matej MartincSenja PollakPublished in: Nat. Lang. Eng. (2019)
Keyphrases
- n gram
- feature vectors
- classification accuracy
- text classification
- feature set
- feature extraction
- feature space
- classification method
- unsupervised feature learning
- rich set
- bag of words
- decision trees
- character n grams
- variable length
- deep learning
- language model
- restricted boltzmann machine
- support vector machine
- text mining
- feature subset
- specific features
- language specific
- unsupervised learning
- image classification
- language modeling
- image features
- deep belief networks
- real world
- inside outside algorithm