Machine Translation of Arabic Interrogative Sentence into English

The present work reports our attempt in developing a bi-lingual Machine Translation (MT) tool in the agriculture domain. The work described here is part of an ongoing research to automate the translation of user interfaces of knowledge-based systems. In particular, we describe the translation of Arabic interrogative sentence into English. In Central Laboratory for Agricultural Expert Systems (CLAES), this tool is found to be essential in developing bilingual (Arabic-to-English) expert systems because both the Arabic and the English versions are needed for development and usage purpose. The tool follows the transfer-based MT approach. A major design goal of this tool is that it can be used as a stand-alone tool and can be very well integrated with a general MT system for Arabic sentence. The paper also describes our experience with the developed MT system and reports results of its application on interrogatives from real agricultural expert systems. Machine translation (MT) is the area of information technology and applied linguistics dealing with the translation of human languages such as English and Arabic. With globalisation and expanding trade, demand for translation is set to grow. Computer technology has been applied in technical translation in order to improve one or both of the following factors (Trujillo Speed: Translation by or with the aid of machines can be faster than manual translation. Cost: Computer aids to translation can reduce the cost per word of a translation. In addition, the use of MT can result in improvements in quality, particularly in the use of consistent terminology within a text or for a particular kind of client. Similarly, the availability of MT makes it ideal for translation of Web pages and other materials on the internet. Translation of Arabic sentences is a difficult task. The difficulty comes from several sources. One is ! ! " #$ %$ & & " ' & $$ words. Another difficulty comes from the sentence structure. The Arabic sentence is complex and syntactically ambiguous due to the frequent usage of grammatical relations, order of words and phrases, conjunctions, and other constructions. Consequently, most of the researches in Arabic MT mainly concentrated on the translation from English to Arabic. An interesting study by (Rafea et al., # ! ( ) ! ! " ! ! " &&! east. A study of ! ! * ( ) & +, & #$$$ study shows the possibility of translating titles of theses and journals from the computer science domain. The translation of an English subset of a knowledge base is described in (El! This tool translates English phrases extracted from the knowledge base that was written in KROL + ! ! . & ( ) ! ! ) & in developing Arabic to English MT system. The big gab between Arabic and English in both lexical and syntactic aspects causes difficulties of building these systems. The lexical gab between Arabic and ! *& & +(! ) ) *& & +Fa *The present work addresses the translation of the Arabic interrogative sentence into English for automating the translation of user interfaces. This includes also the (imperative) form of the verbal sentence that is commonly used for interrogating users of nowadays computer applications. The proposed MT tool described here is part of an ongoing research to automate the translation of user interfaces of knowledge-based systems. In Central Laboratory for Agricultural Expert Systems (CLAES), this tool is found to be essential in developing bilingual (Arabic-to-English) expert systems because both the Arabic and the English versions are needed for development and usage purpose. The next section outlines the overall architecture of the proposed Arabic to English MT system. The following sections describe the main components of the system. In a concluding section, we discuss its application on interrogatives from real agricultural expert systems and present some final remarks. There are three basic approaches being used for developing MT systems that differ in their complexity and sophistication. These approaches are: direct approach, transfer-based approach, and interlingua approach. The current work follows the transfer-based MT approach. There are many factors which make transfer an attractive design for MT (Trujillo : . Many systems are bilingual, or their principal use is for translation in one direction between a limited number of languages. #. Where full multilinguality is required, it is possible to have a hub language into and out of which all translation is done. %. Portions of transfer modules can be shared when closely related languages are involved. For example, an English-Portuguese module may share several transformations with an EnglishSpanish module. The architecture of the transfer-) & ( ) ! , / 0 of the architecture, the arrows indicate the flow of information. Ovals are modules of the system. Rectangles are the linguistic knowledge. The proposed system is based on the transfer approach, with Morphological Analysis and Grammar Rules of Arabic Arabic-toEnglish Transformation Rules Arabic Dic.