On The Development of Statistical Approach Based English to Punjabi Machine Translation System Using Moses Toolkit

Abstract

newline Machine Translation (MT) is a popular topic in the area of artificial intelligence. In this digital era, different communities across the world are connected to each other and are sharing a vast amount of resources. In this kind of digital environment, various natural languages are an obstacle in communication. To remove this barrier, researchers from different countries and big companies are trying to develop machine translation system. Different kinds of approaches have been developed to decode natural languages, for example Rule-Based, Example-Based, Statistical and various Hybrid approaches. Among all these approaches, Statistical-based approach is dominant and popular in the MT research community. The statistical systems yield more accuracy as compared to other approaches, though they need more training data. The resources of Indian languages are less in comparison to European languages, therefore, it is a challenging task to collect parallel corpus for training these statistical models. Efforts have been made to develop parallel phrase corpus in the place of parallel sentence corpus. Collecting parallel phrases is more convenient as compared to parallel sentences.