Semantic processing through distributed representation of Chinese words

摘要

In this paper, by learning the origin of the word distributed representation, knowing the distributed representation is one of the bridges of natural language processing mapping to mathematical calculations. Through the learning distributed representation model: neural network language model, CBOW model and Skip-gram model, the advantages and disadvantages of each model are clarified. Through the experiment, we can understand the mapping relationship between distributed representation calculation and natural language processing, and also clarify the research direction of the next step, that is, the use of distributed representation to sentence, paragraph, article modeling.