Год выпуска: 2011 Автор: Jabar H. Yousif Издательство: LAP Lambert Academic Publishing Страниц: 224 ISBN: 9783844316704
Описание
Arabic language processing has not yet achieved the superiority and automated levels. We aim to utilize and design an automatic POS Tagger for Arabic based supervised neural network techniques, which automatic, accurate, speed, and use least data for training. Review the state of the art of POS tagging methodologies, the characteristics and challenges of Arabic language. The implementation of Artificial Neural Network techniques like MLP, FRNN, SLP, BPN, HOPN, and SRNN are illustrated. The Learning and training algorithms and Genetic Algorithms concepts are presented and implemented. The architecture based NN models and support vector machine is implemented to utilized automatic POS taggers for Arabic applications. The performances of prototype, analyzing the results are assessed using MSE, NMSE, and precision methods. Finally, contributions and future research directions for building automatic information system for Arabic text are discussed and illustrated.
