期刊名称：International Journal of Computer Science Issues
印刷版ISSN：1694-0784
电子版ISSN：1694-0814
出版年度：2012
卷号：9
期号：3
出版社：IJCSI Press
摘要：The speech synthesis is artificial generation of human speech from written texts. For this purpose, adequate algorithms are designed, which then through relevant programs make it possible to synthesize texts to speech. The process of converting text into speech is also known as Text-To-Speech (TTS) system [5]. In this paper are given basic principles to be used when designing a system to synthesize speech in Albanian language from written texts. Currently there are solutions that enable natural speech generation for various world languages. However, unfortunately these are not universal solutions to be used for other languages too, because the volume generated for other languages is incomprehensible and unnatural. For this reason, for every language one should seek solutions that address the specifics of it, always with the aim of generating voice to suit the nature of language. Generating systems that are currently used mainly rely on the use of the concatenation method [6], during which acoustic segments of text files are joined, which are previously digitized and stored as such in a database. For Albanian language, we consider that on the textual part of the database, as basic segments to be used are: the most frequent words, two-letters and letters [4]. However, in a particular part of the database are included various abbreviations, i.e. textual equivalents and their acoustics files, to be used also during the generation of appropriate speech. Whereas, with the aim of synthesizing the various numerical values written in the decimal system, in database were added values, respectively their corresponding sound files, whereby speech is generated for different numbers. The first part of the paper is a brief presentation of the Albanian language [1], respectively of the alphabet used in writing the language and its most frequent words.
