Chhattisgarhi Hindi Machine Translation System Using Rule based Approach

Vikas Pandey, Dr.M.V. Padmavati and Dr. Ramesh Kumar

There are 6,500 languages spoken in the world. So to understand all the languages is a critical problem. This problem is addressed by Machine Translation (MT) which a sub field of Natural Language Processing (NLP). MT helps in converting source to target language. India is a multi-linguistic country where there are 22 major languages in India and over 720 dialects are spoken in India, so conversion of one language to other is a big problem. There are various low resource languages like Bhojpuri, Nimadi, Awadhi and Magahi etc. Chhattisgarhi is also a low resource language because its grammar and dictionary is not available. In our Chhattisgarh state most of the people speak Chhattisgarhi Language. But there are various people who do not speak and write Chhattisgarhi. Even in Government organizations the people face lot of difficulty in writing Chhattisgarhi. Since Chhattisgarhi is a common means of communication in the state so most of the people try to write in Chhattisgarhi. But for those people who do not understand and write Chhattisgarhi it become a difficult task. In order to solve this problem we are attempting to make Chhattisgarhi Hindi machine translation system using rule based approach. Since structural difference between both the languages is less so rule based approach is more appropriate. In this paper, we are presenting a rule based Chhattisgarhi Hindi machine translation system. The system is constructed over 500 rule base and tag set of 30different part of speech tags. The test sentences are taken from various domains including stories and news articles etc. The system achieves an accuracy of 80%.

Volume 11 | 04-Special Issue

Pages: 1099-1108