Identification of Closed Compound Words in Devanagari Scripted and Non-Devanagari Scripted Corpora

Closed compound words and multiword expressions enhance the esthetic quality of any text and are integral components of grammar. In both Hindi and Awadhi, these linguistic constructs are referred to as ‘Samaas’ and play a crucial role in word summarization. This research paper introduces ABD2KT-1.0, an upgraded version of the ABD2KT ‘Samaas’ Finder. The motivation behind this new version was the complexity of rules in the previous iteration. The ABD2KT-1.0 tool serves as a specialized ‘Samaas’ Finder, specifically designed to identify six types of ‘Samaas’. There are six categories of ‘Samaas’ known as ‘Samaas’ Category Identifier (SCI). The examples of SCI are ‘Avyayibhav’, ‘Bahuvrihi’, ‘Dvandva’, ‘Dvigu’, ‘Karmadharya’, and ‘Tatpurusha’ ‘Samaas’. Each SCI follows distinct rules. Upon applying the ABD2KT-1.0 tool to various language words, the authors observed two categories of words: agree words and disagree words. Words adhering to the predetermined ‘Samaas’ rules belong to the agree word category, while those deviating from the rules for any given reason are categorized as disagree words. The tool efficiently handles eight Devanagari and three Non-Devanagari languages. The result stated that the ABD2KT-1.0 recognizes the ‘Samaas’ present in Devanagari languages easily and correctly. For this study authors have taken eight Devanagari languages such as Hindi, Bhojpuri, Dogri, Konkani, Maithili, Marathi, Nepali, and Sanskrit.