الفهرس 
ResultsOnly 14 pages are availabe for public view
 |  | from | 172 |  |  |
| | | from | 172 | | |
Abstract
This thesis presents a realistic technique for machine aided translation system. In this technique, the system dictionary is partitioned into multi-module structure for fast retrieval of arabic features of english words. Each module is accessed through an interface that includes the necessary morphological rules, which directs the search toeard the proper sub-dictionary. Another factor that aids fast retrieval of arabic features of words is the prediction of the word type. then, its subdictionary is accessed to retrieve the corresponding attributes.
The system uses transfer approach that is proceeded in three stages, the source language analysis, the transfer rules between source language and target language and then generation of target language. In the proposed system we use also knowledge-based approach that discards the syntax of the original text, and the output is an interpretation rather than translation. The current system relies on the syntax of the input text, and provides it with semantic information as necessary to solve specific ambiguities. Whereas, the system uses some kind of preference semantics.
The proposed system is able to translate and also adjust nouns, verbs, and adjectives according to their attributes. then, it adds the symptom of arabic word to generate a correct sentence.