This paper is an observation of few attempts under the major area of NLP from the translation perspective especially based on Google Translation, Bing Translation, and Anusaarka, the MT systems for translating Source Language (SL) English to Target Language (TL) Hindi with all the regular/irregular phrases. After the parametric evaluation of these phrase based examples, it’s found that all the systems are having some of basic problems at several stages and the result is bit far from desired output. The following outputs of all the translation systems are under evaluation and whatever translation methodology is used by them till date, has some drawbacks. PBMT (Phrase Based Teaching and Testing Modal) is proved useful in solving these issues. Therefore, this approach can be used for phrase to phrase accurate translation, which will enhance the quality and robustness of statistical as well as Rule-Based system.
Saroj Kumar Jha, Piyush Pratap Singh, Vijay Kumar Kaul. Phrased based T2 model: A review of Google translate, Bing translator & Anusaaraka. International Journal of Advanced Research and Development, Volume 2, Issue 6, 2017, Pages 407-411