| Popis: |
Machine Translation (MT) has proven to be a useful tool for translating one spoken language to another since the mid-,90s. Ultimately, it has helped to bridge the communication gap between people unaware of each other’s languages. A variation of MT, named ‘Sign Language Machine Translation (SLMT),’ enables the translation of spoken text/speech to Sign Language. It helps reduce the ‘hearing’ people’s difficulties in connecting with the hearing-impaired community. Thus, it can aid in the inclusion of hard-hearing in the rest of society. We have developed a system to translate a simple Marathi language sentence to the equivalent Indian Sign Language representation. We have studied the morpho-syntactic characteristics of both languages, and contrastive analysis is discussed here. This study gave us a cue to attempt translation at phrase level from the source language side. We have formulated the rules to extract necessary grammatical knowledge and to utilize it for translation. The detailed approach, experimentation, and results are also addressed in this research work. This automated system can be used at public places like banks, post-offices, railway stations for effective short communications. It can be availed as a self-learning tool to learn Indian Sign Language. The system can be used whenever sign language interpreters are not readily available. |
