Translating English Names to Arabic Using Phonotactic Rules

Translating English Names to Arabic Using Phonotactic Rules ⋆ Faisal Alshuwaier a and Ali Areshey a Computer Research Institute, King Abdulaziz City for Science and Technology P.O.Box 6086, Riyadh 11442, Kingdom of Saudi Arabia {shuwaier, aareshey}@kacst.edu.sa Abstract. With the increasing numbers of the arrivals to the Arabian countries; it is neces- sary to use Arabic language in writing names on official documents. Because of the difference in writing English names in Arabic language, many methods have been spread for translating English names which led to first, duplication in every single English name written in Arabic and this will lead to negative effects security and property rights. The second difference is that even if the transliteration of Arabic names is standardized, it is difficult for a layperson to implement it. This paper is to provide algorithms based on some rules that can be used in programming a system to transliterate English names automatically. The system uses only plan for Translate English Names to Arabic, and that can be processed and printed easily. Moreover, the Translated names can be read and recognized by ordinary people. Keywords: Automatic Translation, Phonotactic Rules, Phonemes, Arabic, English, Dia- critized Arabic Phone. 1 Introduction With the rapid increase of the published information, it is difficult to rely on human translation to translate such information from one language to another and translation of appropriate names is generally realized as a significant issue in many multi-lingual text and speech processing applica- tions. In our work, a set of hand-crafted transformations for locally editing the phonemic spelling of an English word to conform to rules of Arabic syllabification are used to seed a transformation- based learning algorithm. The algorithm examines some data and learns the proper sequence of application of the transformations to convert an English phoneme sequence to a Arabic syllable sequence. Our paper describes a data driven counterpart to this technique, in which a cascade model is used to go from English names to Arabic transliteration. The system uses only plan for Translate English Names to Arabic, and that can be processed and printed easily. Moreover, the Translated names can be read and recognized by ordinary people. This paper also discusses the problem background and provides solutions to Arabize almost all English Names. In section 2, we briefly highlight some related works. In section 3, we introduce description of translation system. In section 4, some specific phonotactic rules are shown. In section 5, the evaluation experiment for auto transliteration is presented. Finally, in Section 6, we draw some concluding remarks. 2 Related Works Transliteration has been of interest to researchers in automatic lexicon creation and cross-language information retrieval involving radical differences in writing systems, such as English/Arabic, English/Korean, English/ Chinese and Japanese/English. (Knight and Graehl, 1998) have relied on the probabilistic mappings between English phones and Japanese phones generated. (Mansur ⋆ The authors would like to acknowledge the reviewers for their valuable comments, which contributed to the clarity of the paper and in particular for their suggestions for …

Content truncated to 500 words