Laba Kr. Thakuria et al Int. Journal of Engineering Research and Applications www.ijera.com ISSN : 2248-9622, Vol. 4, Issue 2( Version 1), February 2014, pp.446-450 www.ijera.com 446 | Page Automatic Syllabification Rules for ASSAMESE Language Laba Kr. Thakuria 1 , Prof. P.H. Talukdar 2 Department of Instrumentation & USIC, Gauhati University Guwahati, India Department of Instrumentation & USIC, Gauhati University Guwahati, India Abstract For unit selection based text-to-speech system, syllabification acts as a backbone. Based on different structures of different languages syllabification rules are also varies. The purpose of this study is to examine and analyse the syllabification rules for Assamese language. Imparting education and training, preferably, in the local/regional language is urgently necessary in today’s context in order to maintain social harmony and homogeneity. Language heterogeneity is a global problem in bringing all the benefits of Information Technology (IT) to our doorsteps. Syllabification rules are implemented into an algorithm which later can be integrated into a text-to-speech system. The analysis of these rules has been taken using 10000 phonetically rich words which reports to produce a comparable result of 99% accuracy as compared to manual syllabification. Keywords: Assamese language, diphthongs, phonemes, syllable, text-to-speech. I. Introduction Syllable is a unit of sound which is larger than phoneme and smaller than a word [5]. Syllabification algorithms are mainly used in text-to- speech (TTS) systems in producing natural sounding speech, and in speech recognizers in detecting out-of- vocabulary words[4]. Syllable forms as a gap between a phonemes and words [1]. Various attempts have been made to define a syllable earlier. According to phonetics it is defined with respect to its articulation whereas in phonology it is simply termed as a sequence of phonemes [3].When a word is broken down into its syllables the process is called syllabification. As we humans also syllabify a word, as far as possible before speaking if possible and phonemic segmentation if not, text-to-speech systems usually use syllable based approach as a basic unit [6]. A syllable based text-to-speech system performs always better than a phoneme level approach in terms of naturalness and easy in boundary analysis.In this study there is an honest endeavour to syllabify ASSAMESE language as well as implemented an algorithm which further automatically divides words into its constituent syllables.It is the aim of the proposed paper to study the phonetic variations of Assamese language to develop syllabification rules for Assamese language. the algorithm was tested using 5000 phonetically rich words. The degree of accuracy of the algorithm is 99%. The research paper is organized as Assamese Language and its Phonological structure. It also describes the Family tree of the Assamese Language. Then it describes the Syllable structure and its algorithm with rules. This section also includes its working methodology. II. Assamese Language And Its Phonological Structure Assamese is an Indo-Aryan language spoken by the Assamese people in general. The mixed Aryan culture and the mongoloid culture gave birth to a new culture. So, every community from this region always exhibits their indigenous culture with diversity. It is the link language for the people living in Assam and adjoining states of Arunachal Pradesh, Meghalaya, Nagaland etc. This language has come from Sanskrit as its offshoot, through different stages of development. Fig-1: Proto Indo-European Family Tree The ASSAMESE phonemic inventory consists of eight vowels, ten diphthongs, twenty-one consonants and two semi vowels [3]. The ASSAMESE vowels and consonants are shown in the tables below. RESEARCH ARTICLE OPEN ACCESS