Research on Feature Extraction and Tendency Analysis of Chinese Language Sentimental Expression in Network Context

The current network language only considers the expression of its basic meaning, and lacks in-depth analysis of sentimental characteristics and tendencies. For this reason, this paper investigates the sentimental analysis methods of network language based on the Chinese text information of network social media platform. Through the training of the existing data sets, we complete the data cleaning and feature word extraction. Then, a multi-dimensional fine-grained sentiment classification algorithm is designed by using neural network to reduce the dimension of feature vector. Finally, the structured linguistic features are used to optimize the sentiment score step by step, and the text sentiment classification is implemented, and a tendency analysis model considering the polarity of sentiment is further established. The validity of the model is verified by the real data collected in the network. The results show that the method has a significant advantage in Chinese data sets, and the recognition rate is higher than that of similar sentiment analysis methods.