2020/10/11 by Khalid Alnajjar, Mika Hämäläinen, Alnajjar, Khalid +5
Computer Science · Social Sciences · #Natural Language Processing Techniques #Historical and Linguistic Studies #Mathematics, Computing, and Information Processing
paper · pdf · doi:10.48550/arxiv.2010.05269
This study uses a character level neural machine translation approach trained on a long short-term memory-based bi-directional recurrent neural network architecture for diacritization of Medieval Arabic. The results improve from the online tool used as a baseline. A diacritization model have been published openly through an easy to use Python package available on PyPi and Zenodo. We have found that context size should be considered when optimizing a feasible prediction model.