Disambiguating Arabic Words According to Their Historical Appearance in the Document Based on Recurrent Neural Networks.
Rim LaatarChafik AloulouLamia Hadrich BelguithPublished in: ACM Trans. Asian Low Resour. Lang. Inf. Process. (2020)
Keyphrases
- recurrent neural networks
- historical documents
- handwritten documents
- printed documents
- arabic language
- text documents
- printed text
- unknown words
- word sense
- keywords
- document images
- document representation
- handwriting recognition
- related words
- feed forward
- index terms
- word co occurrence
- arabic documents
- neural network
- arabic text
- keyword extraction
- historical manuscripts
- reservoir computing
- latent topics
- recurrent networks
- word recognition
- topic hierarchy
- word spotting
- compound words
- echo state networks
- feedforward neural networks
- document collections
- document level
- information retrieval
- document analysis
- web documents
- artificial neural networks
- co occurrence
- text corpus
- nonlinear dynamic systems
- document content
- information retrieval systems
- word segmentation
- word sense disambiguation
- neural model
- word level
- multiword
- text mining
- tf idf
- vector space model
- character recognition
- bag of words
- n gram
- scanned documents
- cascade correlation
- text classification
- handwritten words
- topic models
- pose estimation
- query expansion
- text summarization
- optical character recognition
- information extraction
- artificial intelligence
- machine learning