NLP ( Natural Language Processing ) can be used to communicate with computers by means of intelligence methods in a natural language. In other words, it is very useful for classification and analysis. Although there are many natural language syntax analyzers that existed before, it still remains to fulfill the requirements for analyzing English text for English to English Machine Translation. In this paper, we have proposed a chunk based analyzer for English to English Machine Translation System. The keyword search Technique is one of the techniques of Natural Language Understanding. A simple keyword based matching technique is used for classification. Domain-specific dictionaries of keywords are used to reduce the dimensionality of feature space. The output result is that the word matched to a known word will be replaced in the given sentence, which will not be changed the original meaning.