Research on Recommendation Framework for Personalized Push of Cultural and Creative Products in Tourist Cities Based on Multi-Level Computational User Profiling

Currently, the development of cultural tourism has become a new trend of urban development, and how to use modern technology to realize the innovative development of urban cultural tourism has become a key issue to be considered in the process of urban construction.The research combines the Web domain ontology to construct a multi-level user portrait master model, which mainly includes four sub-models: retailer static attribute vector model, retailer domain dimension model, retailer marketing ability model and retailer social dimension model.The FCM algorithm based on the improved AP algorithm is utilized to cluster the user portraits, and the user portrait clusters obtained by the method studied in this paper perform well with an average number of iterations and an average time consumed of 21.3 and 60.35 compared with the traditional K-Means algorithm, the improved K-Means algorithm, and the traditional FCM algorithm, respectively.Then a personalized recommendation method for tourism products based on MAGFM is proposed, which achieves Top-N recommendation of tourism products by calculating the total interest value of users and the comprehensive similarity of tourism products.And test and analyze in the tourism e-commerce platform, the results show that the recommendation algorithm proposed in this paper has higher effectiveness compared with the traditional recommendation algorithm.Finally, the research content builds a personalized recommendation system for tourism cultural and creative products.