Research on Dependency Parsing based on Optimized Neural Networks

In view of the complex structure of Chinese sentence patterns, it is difficult to accurately extract the feature information of dependency structure in dependency analysis, which directly affects the accuracy of correlation analysis. In this paper, based on stanfordnlp, the sentence dependency structure cannot be accurately extracted in Chinese dependency analysis. BiLSTM(Bi-directional Long short-term Memory) is used to solve the problem of Long distance dependent feature extraction, and the context feature information is further extracted by combining conditional random field (CRF).Then, a BiLSTM-CRF optimized neural network for Chinese dependency analysis is proposed, and the experimental results are verified. The experimental results show that it is effective to improve the accuracy of Chinese dependency analysis.