Artistic creation in the network era, examined in terms of artistic style, can be roughly divided into several important stages in time, such as 1985-1995, 1996-2005, 2006-2015, and 2016-present. This paper integrates pedagogical knowledge distillation and an artistic style feature extraction model to investigate the change in artistic creation style in the network era. Using global features, the color histogram of the art image region is extracted by calculating the color features on H, S, and V from both color and texture features. The color and texture feature vectors are normalized, assuming that they conform to a Gaussian distribution. Quantify the art style features using an adaptive weighted Gram matrix to improve the accuracy of art image style classification. Joint art knowledge distillation model for dissemination of knowledge of the extracted art style features. Plotting the scatter of color chromaticity values L*-b* of all painting samples reveals that the luminance L* of all paintings fluctuates between 21.38 and 79.94, the red/greenness value a* fluctuates between -8.57 and 1.7, and the yellow/blueness value b* fluctuates between 1.67 and 16.21. The temporal development of the color characteristics of art styles follows a clear pattern. In the comparison of texture eigenvalues under variable parameters, the moment of inertia of the painting images in the period 1985-1995 is raised from 147.698 to 615.965 after noise addition, and there are different degrees of growth in the moment of inertia in the other periods, but with different magnitudes of growth, and there are obvious differences.