NLP Tools and Resources to Revitalize Potentially Vulnerable Low Resourced Kudmali Language: A Brief Overview
摘要
The Kudmali language, an official language of Government of West Bengal, like other indigenous and lesser-spoken languages in and around the Chota Nagpur Plateau, West Bengal, Jharkhand and Odisha is under threat. There are various factors of endangerment of this language such as globalization, urbanization and the dominance of more widely spoken neighboring languages with code mixing of those. According to UNESCO Kudmali Language falls under the category potentially vulnerable. Languages carry our culture, cumulative knowledge and values are part of the identities of any communities. The importance of indigenous languages is now well-established. NEP 2020 has also focused on the promotion of all Indian languages and mother tongues irrespective of their status. The objective of this paper is to highlight the way to save the language and to encourage NLP researchers to delve into its study and preservation. This paper gives a brief overview of the current state-of-the-art NLP methods for documenting and making different NLP tools and resources for preserving the language by improving its digital presence. This study is structured around several key objectives: the compilation and annotation of a comprehensive Kudmali text and speech corpus, explaining different linguistic features of Kudmali language which will attract the NLP researchers to work on and explaining the techniques to make different NLP applications to support Kudmali language including tokenization, part-of-speech tagging, named entity recognition, and machine translation, Automatic Speech Recognition model.