Innovation and Practice of Ideological and Political Education Communication Mode Driven by Artificial Intelligence

The combination of “artificial intelligence” and civic education has become an important innovative direction and dissemination channel for students’ civic education. In this paper, we construct a recommendation algorithm based on graph neural network for Civics and Political Science, which aggregates the information of neighboring nodes based on graph neural network, and uses the attention mechanism to distinguish the different effects of different strong and weak relationship nodes on the target node, so as to improve the recommendation performance of the algorithm. An exclusive optimal interpretation vector is learned for each interaction and social relationship, and the multifaceted preferences inherent in the audience for Civics content as well as the multifaceted influences of the audience on each other are mined, and the model prediction module utilizes the hidden vectors of the student nodes and the hidden vectors of the Civics content nodes obtained from the previous module to obtain the final predicted student ratings of Civic content through the multilayer perceptual machine. The performance of the designed algorithm is examined on datasets such as Gowalla and Yelp2018, and it is found that the prediction results of this paper’s model are greatly improved over the baseline model, because the social network information is taken from real social information, and the algorithm capture will be more accurate after more information is captured from it. In the application of Civic Communication, the model in this paper can grasp the propagation cycle law of high retweeted Civic content, which is in line with the law of network communication, and helps to carry out Civic Public Opinion Control and Civic Education and Teaching.