Machine Intelligence for Language Translation from Kashmiri to English

Machine translation (MT) is an emerging research as well as application area in the contemporary world. It is receiving significant attention from academia, industry and corporate houses. A wide range of translation techniques are being applied either individually or in combination for machine translation of different languages across the globe. However, there are still many languages that are either completely missing or poorly visible on the machine translation map. The Kashmiri language is one such language where very little or negligible work has been done related to its machine translation. This paper aims to present a Kashmiri-to-English Machine Translation System and highlight various features of the Kashmiri language. The system is based on machine intelligence having the ability to learn various translation rules from the translated set of input sentences, using Long Short-term Memory (LSTM) architecture for deep sequence learning. The paper also discusses various challenges related to machine translation from Kashmiri to English or other languages. The work presented in this paper is the first of its nature and can serve as a bedrock for research community interested to work on machine translation of Kashmiri language.