Research on the Quantification and Multi-dimensional Evaluation Model of Ideological Education Information Dissemination Effect Based on Deep Learning in the Post-Pandemic Era

As the epidemic subsides, the fast development of the digital media and the wide application of online platforms have become huge changes in the dissemination of ideological and educational information, resulting in the urgent need for a more effective and accurate evaluation method. In this paper, a deep learning-based model quantifying and multi-dimensional evaluation the influence of information dissemination to ideological education is introduced that uses Convolutional Neural Network (CNN) as the main architecture, and optimizes it through advanced small methods. To further optimize the CNN architecture, we introduce advanced small methods such as dynamic adjustment of convolution kernel sizes and shapes through adaptive convolutional layers. This allows the model to learn more details about the features in education contents. In addition, we present a temporal convolution with spatial feature extraction that simultaneously extracts time-sensitive spatio-temporal temporal patterns of audience perception. The resulting spatio-temporal evolution is derived from multi-dimensional visualizations, which includes sentiment, interaction, and content engagement over time. The experimental results demonstrate that the proposed model enables better accuracy in prediction and evaluation of communication effect than the traditional methods, and can effectively provide guidance to the optimization and adjustment of ideological education content.