Fine-grained Emotional Analysis and Intelligent Identification of Electric Power Public Opinion Based on Attention Mechanism

In the Internet age, information explodes exponentially. As an electric power industry closely related to national life, network public opinion in the electric power field also spreads rapidly. How to fully tap the public opinion information of electric power to promote the improvement of electric power service is an important research topic. According to the research status of power public opinion, this paper defines the problem in fine-grained emotion analysis, collects the corresponding data, and defines the evaluation aspects and evaluation levels. In terms of solution strategy, according to the dimension of the problem, this paper adopts a two-way LSTM model integrating attention mechanism. The experimental results show that this method is effective in the fine-grained emotion analysis task of power public opinion samples.