In today’s deepening globalization, arts and crafts teaching is facing unprecedented challenges and opportunities. Cross-cultural art exchange is not only a trend, but also a necessity. The confidence and support of association rules affect the quality of mining, in order to improve the mining efficiency, this paper proposes an algorithm based on improved gravitational search mixed with particle swarm (GSA-PSO) for association rule mining. When mining association rules, support, confidence and boost are selected as the evaluation criteria of association rules and the optimal rules are selected using Pareto method. Through simulation tests on nine datasets of different sizes, the experiments prove that the algorithm proposed in this paper has a significant advantage over the other five algorithms in terms of the quality of the rules obtained. Using the dataset of Chinese and Western painting images for analysis, based on the Sankey diagram composed of the association relationship between WesternChinese paintings, the adaptive threshold mining results of Western-Chinese painting types such as {spatial perspective-artistic conception} are strong association rules with the maximum enhancement greater than 1, but the enhancement for the fixed threshold {300m, 1h} is less than 1.
