In response to the problem of low recommendation accuracy and inability to meet user preferences in existing methods, the Jin Asia Pacific SOM-3568-SMARC industrial grade core board is used as the hardware foundation, supplemented by storage modules to achieve data processing. At the software level, an interactive trust network model based on the topological characteristics of user interaction networks has been constructed, which deeply mines and integrates multi-source information between users, accurately quantifies the strength of interactive trust, and then solves the objective function through optimization algorithms to achieve personalized and high-precision recommendation of intangible cultural heritage information. In the test results, the F1 Score of the designed system remained stable at 0.90 or above, indicating high recommendation accuracy and higher user satisfaction, indicating that its recommendation results can meet user preferences.