Labeling of the Arabic Words Hammad BALLAOUI El Habib BEN LAHMAR Hassan II University - Department of Math and Computing, Faculty of Science, Chouaib Doukkali University, El jadida, Morocco. habahammad@gmail.com El Khadir LAMRANI Hassan II University – Mohammedia -Casablanca, Ben M'sik Facultyof Science, Information Technology and Modeling Laboratory, Casablanca, Morocco Mohammedia - Casablanca, Ben M'sik Faculty of Science, Information Technology and Modeling Laboratory, Casablanca, Morocco. khadir.lamrani@gmail.com Abstract In this article we present our works on the morphosyntactic labeling and more particularly the determination of an Arabic word to the base of its affixes (prefixes and suffixes), of its particles of attributions and in contexts. Our approach bases itself on an Automat system in contextual words. The objective contextual words. The objective of h.benlahmer@gmail.com Nasser LABANI Department of Math and Computing, Faculty of Science, Chouaib Doukkali University, El jadida, Morocco. Deterministic Finished State (AEFD). This proposed technique allows to analyze the words in two stages. The first stage shows that the system is going to handle every word entering as a noun, a verb or a particle to be attributed by one of the latter. The second stage constitutes a graphical representation of the set of the units found previously to determine the number and time do the words belong? the approach presented here is to associate a label with every lexical unity and to deduce , in any final state of the automaton, to which category, kind, 1. Introduction The automatic morphosyntactic labeling of an Arabic text is an essential stage of its analysis, and useful for numerous applications of the automatic treatment of Arabic language (TALA). Thanks to the advent of Internet and search engines, where the problem is not anymore the access to the information [1], the Arabic language natural processing has known these last years (since 2000, so far) a true ascent whether it is on various scientists plans (the design of the Latin-Arabic or Arabic-Latin scientific dictionaries and the mechanisms which allow to handle concepts in ontology ...) ,or social factors (the encouragement of the Arabic researchers to develop and improve their searches …) or economic (sale of products and computing software …); all that is done by the emergence of a very important number of the technical inventions such as the automatic translators of texts, spell- checkers of errors, and automatic generators of summaries … etc. [2]. But even with these technical developments, there are still several difficulties in the automatic treatment bound by the characteristics of the Arabic language itself, in a point which certain researchers asserted that there is no complete system for the labeling of Arabic text [3]. Today, this linguistic space (TALA) entails much more effort to reach a successful result, because the former attempts do not often manage to buckle the majority of the linguistic phenomena in Arabic, and besides, the performances of search remain defective especially in the domain of the computerization of the Arabic language. But the only linguistic and IT challenge which is the main stumbling block and which often hampers the researchers for the automatic treatment of a natural language (TALN) is the problem of the ambiguity. This problem strongly increased in Arabic language unlike the other natural languages such as English and French; this complexity shows itself under Hammad Ballaoui et al, Int.J.Computer Technology & Applications,Vol 6 (5),705-711 IJCTA | Sept-Oct 2015 Available online@www.ijcta.com 705 ISSN:2229-6093