Abstract: 【Purpose/significance】Exploring the patterns and characteristics of interdisciplinary knowledge integration is of great sig⁃ nificance for promoting disciplinary development and complex problem research. 【Method/process】This article constructs interdisci⁃ plinary networks in different countries and proposes a knowledge integration difference recognition method that combines Graph Ker⁃ nel and LDA topic models. The article analyzes the attribute features, similarity, and differences of interdisciplinary networks.【 Result/ conclusion】The results indicate that this method can effectively identify the characteristics and differences of different countries in in⁃ terdisciplinary knowledge integration. There are both similarities and differences in interdisciplinary preferences among different coun⁃ tries; There are also differences in the tendency of knowledge fusion in the same interdisciplinary direction among countries【. Innova⁃ tion/limitation】The methods and related conclusions proposed in the study are helpful in promoting interdisciplinary research, but due to space limitations, only the main interdisciplinary directions in the field were analyzed.