Document Type : Research Paper

Authors

• Abeer K.Al-Mashhsdany ¹
• Abdulwadood K. Al-Mashhadany ²
• Waleed K. Al-Mashhadany ³

¹ University of Al-Nahrain - College of Science

² Iraqi Ministry of Health, Baghdad

³ Iraqi Ministry of Education, Baghdad

Abstract

Many application programs require knowledge from the user. In such programs, number of problems may occur. Problems related to natural language understanding such as; typos, duplication, similarity, and inconsistent knowledge. So, the analysis of text is very important in such programs. A previous work was done to solve problems of natural language understanding. That work build Arabic analyzer and merge it with a shell hybrid system. This work attempts to build a general Arabic Analyzer system, so that it could be merged with other programs to solve the natural language understanding problems. This work attempts to correct the previous Arabic analyzer, so its dictionary will be smaller. Morphology in this work merges root-based with stem-based approach. This merging forces this work because it gains the advantages of the two approaches. Root-based approach gives the advantage of smaller size. Stem-based approach gives the advantage of passing the problem of irregular cases. This work is compared with other two valued techniques. Results of the comparison show that this work has a lot of advantages over the previous techniques.

Keywords

• root
• stem
• based approach
• Shell Systems
• Word Analysis
• Arabic Morphological Rules
• Synonyms and antonyms words