The IUCL+ System: Word-Level Language Identification via Extended Markov Models.
Levi KingEric BaucomTimur GilmanovSandra KüblerDan WhyattWolfgang MaierPaul RodriguesPublished in: CodeSwitch@EMNLP (2014)
Keyphrases
- markov models
- language identification
- word level
- document images
- markov model
- maximum entropy
- higher order
- hidden markov models
- document analysis
- language independent
- n gram
- conditional random fields
- speaker identification
- markov chain
- optical character recognition
- sentence level
- machine translation
- word recognition
- visual features
- information retrieval