NLI-GSC: A Natural Language Interface for Generating SourceCode

There are many different programming languages and each language has its own structure or way of writing the code, it becomes difficult to learn frequently switch between languages. Due this reason, a person working with multiple needs look at documentations which costs time effort. In past few years, there have been significant increase in amount papers published on topic, providing unique solution problem. Many these based applying NLP concepts configuration get desired results. Some used AI along train system generate source-code specific language, some trained directly without pre-processing dataset NLP. All face two problems: lack proper for particular application paper can convent natural into only one specified source-code. This proposed shows that independent is feasible alternate having full knowledge about language. The uses Natural Lan-guage Processing convert Language language-independent pseudo code using custom Named Entity Recognition save XML (eXtensible Markup Language) format an intermediate step. Then, traditional programming, converts generated language-dependent paper, another novel method create from scratch predefined filled keywords creating combination training dataset.