Automatic Phonetization-based Statistical Linguistic Study of Standard Arabic

Statistical studies based on automatic phonetic transcription of Standard Arabic texts are rare, and even though studies have been performed, they have been done only on one level – phoneme or syllable – and the results cannot be generalized on the language as a whole. In this paper we automatically derived accurate statistical information about phonemes, allophones, syllables, and allosyllables in Standard Arabic. A corpus of more than 5 million words, including words and sentences from both Classical Arabic and Modern Standard Arabic, has been prepared and preprocessed. We developed a software package to accomplish a rule-based automatic transcription from written Standard Arabic text to the corresponding linguistic units at four levels: phoneme, allophone, syllable, and allosyllable. After testing the software on four corpora including more than 57000 vocabulary words, and achieving a very high accuracy (> 99 %) on the four levels, we used this software as a reliable tool for the automatic transcription of the corpus used in this paper and evaluated the following: 1) the vocabulary phonemes, allophones, syllables, and allosyllables with their specific percentages in Standard Arabic. 2) the best curve equation from the distribution of phonemes, allophones, syllables, and allosyllables normalized frequencies. 3) important statistical information, such as percentage of consonants and vowels, percentage of the consonants classified by the place and way of articulation, the transition probability matrix between phonemes, and percentages of syllables according to the type of syllable, etc. Fadi Sindran, Firas Mualla, Tino Haderlein, Khaled Daqrouq & Elmar Nöth International Journal of Computational Linguistics (IJCL), Volume (7) : Issue (2) : 2016 39