Natural Language Understanding by Natural Language Programming

Abstract

The main purpose of this paper is to introduce an advanced programming technique "dynamically run" in Natural Language Programming. And I will show how this technique is used in natural language understanding. This technique not only greatly simplifies the source codes, but also reduces the machine code loading time complexity initially. Meanwhile, I design a parsing method that scans the sentence once and which uses additional list and stack possibly for the context free structure. To get the syntax and semantics of the sentence, each word is run as a function which is defined inside one file already. And the file of this function is loaded right before it is run dynamically. Lastly, to complete the understanding of natural language sentences, I introduce the concepts of knowledge representation and information flow.