Part of speech tagging for Arabic text based radial basis function

Abstract

The progress of peoples and their advancement linked to the extent of their keenness to spread their civilization and scientific progress in their language and tongue. Language is the pot of civilization without a doubt. The Arabic language is one of the most spoken languages in the world, with more than 467 million people speaking it. The Arabic language has succeeded in playing a unique civilized role, making it the leader of civilization and knowledge at the level of science for centuries in a row. In the current era, because of the proliferation of modern technologies, advanced technology and software that take the foreign languages its own platform, the role of the Arabic language has declined in influence, and it has become imprisoned in universities, schools and the role of science that only its people can learn. However, at the present, it is time to gather its forces to meet the requirements of the present and the future in the technical, civilizational, and epistemological field. The aim of this work is to detect the Part of Speech (POS), in which to simplify the rules of the Arabic language to ensure its vitality by keeping pace with the language of technical developments. In addition, to spread Arabic language sciences through the application resulting from this reality via the Internet, which makes it the greatest influence in simplifying its own rules for non-native speakers and the desire to study and learn its arts. The proposed model consists of three phases, that starting by the phase of reading the text and segments it into sentences and words, followed by the phase of analyzing and tagging each word inside the text. Finally, the syntax analyzer (Parser) that used to determine the (POS) will built by using the Radial Basis Function network (RBF) which is a type of Artificial Neural Network (ANN). RBF simply contains three layers (layer that is responsible for entering data, the hidden layer in which processing is done, and adjusting the network weights to perform their role in matching the inputs and outputs, and the output layer).