Reform path and empirical exploration of visual communication design education based on graph neural network

With the continuous development of visual communication design (VCD), its design concept is also constantly evolving. Traditional graphic design education is no longer meeting the needs of contemporary design education, making reform essential to improve the teaching quality of VCD. Based on graph neural network (GNN), this paper is taking the VCD course as an example to explore effective paths for VCD education reform. The experimental subjects are being divided into two groups: the experimental group is adopting the VCD course reform teaching mode based on GNN, while the control group is using the traditional teaching mode. Through comparative analysis of the two groups' data, it is being found that the attendance rate of students in the experimental group is reaching 97%, which is 9% higher than the 88% attendance rate in the control group. Similarly, when comparing the test scores of the two groups, 5% of the students in the experimental group are scoring above 95 points, 69% are scoring between 94 and 80 points, and only 2% are scoring below 60. In contrast, in the traditional teaching mode, 56% are scoring between 94 and 80 points, 37% are scoring between 79 and 60 points, and 5% are scoring below 60. These results are indicating that the learning level of students in the GNN-based VCD course reform teaching mode is improving significantly. The findings are demonstrating that the GNN-based VCD teaching reform is having a positive impact on students' learning efficiency and innovation ability, and is providing new ideas for the ongoing reform of VCD course teaching.