The Entropy of Words - Learnability and Expressivity across More than 1000 Languages.
Christian BentzDimitrios AlikaniotisMichael CysouwRamon Ferrer-i-CanchoPublished in: Entropy (2017)
Keyphrases
- arabic language
- pattern languages
- language specific
- word forms
- expressive power
- language independent
- keywords
- mutual information
- learning algorithm
- related words
- information theory
- word order
- information theoretic
- multilingual documents
- n gram
- arabic documents
- regular languages
- grammatical inference
- word sense disambiguation
- language identification
- cross lingual
- indian languages
- text documents
- finite automata
- parallel corpora
- statistical machine translation
- word pairs
- boolean functions
- syntactic categories
- topic models
- learning from positive data