A Geometric Method to Obtain the Generation Probability of a Sentence

Abstract

"How to generate a sentence" is the most critical and difficult problem in all the natural language processing technologies. In this paper, we present a new approach to explain the generation process of a sentence from the perspective of mathematics. Our method is based on the premise that in our brain a sentence is a part of a word network which is formed by many word nodes. Experiments show that the probability of the entire sentence can be obtained by the probabilities of single words and the probabilities of the co-occurrence of word pairs, which indicate that human use the synthesis method to generate a sentence.