Abstract: There are vast develop in Natural Language Processing(NLP) tools such as Part Of Speech (POS) and morphology analysis, but the IE system for an Arabic texts are still not rising to comply with such progress and it is not available till now. The paper clarifies the design and implementation of a suitable information extraction system that automates input feeding of an Arabic text to derive an output template as a convenient extraction of important information. An Arabic POS tagger based on multilayer perceptron neural network (MLP) is used to achieve high accuracy and scouring for the tokenization stage. In addition, an Arabic semantic parser is suggested and addressed. The system consists of two stages, the first one is preprocessing stage, which has used to implement and model the document analysis and tokenization. The other stage is the processing stage, which has used to accomplish the morphology analysis, Pos tagger, semantic tagger and name entity extraction.