A Hypothesis of Neural Network about Chinese Language Understanding

To explain the human understanding and constructing of Chinese sentences, a hierarchical neural network hypothesis was proposed from the point of view of cognitive science, psychology, and neural physiology. The cognition of Chinese characters, words and sentences was accomplished by corresponding neuron regions, associated with which other neuron regions implied the meaning (or semantics) of these language objects. A series of activations between neuron regions can explain how human brain understands and constructs a sentence.