Study on Turning Arabic Text into Spoken Words

• Main Authors: Abdulwahhab Shareef, Riyadh Mahmoud
• Format: Article
• Language: Arabic
• Published: Mosul University 2021-06-01
• Series: Al-Rafidain Journal of Computer Sciences and Mathematics
• Subjects: sound recording; analysis; synthesis; overlap; smoothing speech
• Online Access: https://csmj.mosuljournals.com/article_168269_34025355b9877adb4458942531247cee.pdf

Language was a means of communication between members of a single community, so that some of them would express to each other their ideas and thoughts, and a common characteristic of that society was defined, and the origin of the word was derived from (rhetoric, idle), if he repealed the matter and spoke about it, and the ancients knew it that language: (It is what expresses It contains all the people for their belongings). The ancients went in their definition of a language to the characteristics of their language with which they communicate, without referring to the Arabic language as a language among the living languages ​​that are circulating among them. Arabs and Arabs (whoever inhabited the country and its island, or uttered the language of its people), and to it returns the percentage of the Arabic language, which is one of the Semitic languages. Which spread in the Arabian Peninsula, and the writing came to denote what is in the minds of notables. This research presents a computer application that depends on human input to pronounce the Arabic letters The system consists of two phases, the first stage is the axis of creating a database for Arabic language characters and their storage locations, as well as the type of formulas for those letters when the initial processing was performed. The second stage is the process of comparing the entered letter from the text with the corresponding sound and placing it in a storage so that we can then process it. In the practical part of the research, we used a comparison between the results of four methods to obtain the least possible execution time with the least pauses in speech, which are the combinative method, the smoothing method, the method of nesting speech, and a hybrid method between smoothing and interfering together. We start by entering in the input text box and using the SpellLetter function, which we use for the purpose of processing and pronunciation, where the input is of three types, either it is a char, or it is an array of characters (String), or it is numbers between zero and nine Num), and in the case that the entry was not For these three types, the entry is wrong, for example, non-Arabic letters or special symbols, for example.,And if the entry was correct, each letter is taken with its accent. This means that we take two positions each time. Using Matlab (R 2018a) to build the proposed system and it was implemented using a computer. Portable running under the environment of the operating system (Microsoft Windows 10).