Author(s)

NanFang Jian

Corresponding Author

NanFang Jian

Abstract

Dependency parsing has been the mainstream of Syntactic parsing in NLP community. With the rapid development of deep neural network, more researches tend to utilize the neural network for dependency parsing, all these efforts has improved the accuracy of dependency parsing greatly. In this paper, we first introduce the theory of dependency parsing briefly. Then the main process of dependency parsing is analyzed in detail, including feature learning, training and decoding process. Lastly, we summarize the current problems and challenges during dependency parsing research, as well as discussing the further development.

Keywords

Graph-based dependency parsing, Deep neural network, Recurrent neural network, Word embedding.