Implementation Approach of Indian Language Gujarati Grammar's Concept “sandhi” using the Concepts of Rule-based NLP

Abstract

The term ‘language’ in NLP has to be understood as natural languages like Gujarati, Hindi, English etc., which we use in daily life to communicate. Most of the NLP research has been centered on English & other European Languages. NLP research concerning the Indian language like Gujarati is commenced in the last few years. The centre of attention of this paper is to demonstrate the road map of implementation of Gujarati grammar's concept “sandhi ”. In our words sandhi is a word segmentation process & it is present in most of the South Asian language, such as Devnagri, Sanskrit, Hindi, and Gujarati & even in Chinese & Thai languages.” Sandhi leads to phonetic transformation at word boundaries of a written chunk (small part), and the sounds at the end of word join together to form a single chunk of the character sequence.” Our main spotlight is on rule-based implementation of “sandhi”. Similar to every Indian scripting language Gujarati language (Grammar) also has its own specified rules of composition for combining the consonants, vowels and modifiers. We have identified certain rules by which we accomplish the practical implementation of “sandhi ”. There are many sandhi rules available, each denoting a unique combination of phonetic transformations, documented in the grammatical tradition of Gujarati. The Sandhi does not make any syntactic or semantic changes to the words implicated. Sandhi is an elective operation that depends only on the alertness of the writer.