Research On Collaborative Forecasting Of Smart Tourism Supply Chain Based On CPFR

The 14 th Five-Year Plan calls for "in-depth development of mass tourism and smart tourism,innovation of tourism product system,improvement of tourism consumption experience","perfect tourism infrastructure and distribution system,and strengthen the construction of smart scenic spots",as well as the use of cloud computing,5G,big data,Io T and other modern information technologies to meet people’s demand for diversified and personalized travel services.It makes smart tourism become the inevitable trend of tourism development.From the perspective of smart tourism supply chain,this paper firstly analyzes the current smart tourism and its forecasting status by consulting data,and explores the existing problems in smart tourism supply chain forecasting.Secondly,the reasons for the problems of smart tourism supply chain prediction are analyzed from the perspectives of smart tourism scenic spot construction and smart tourism supply chain.Then,based on CPFR theory,a collaborative model composed of collaborative mechanism and collaborative module of smart tourism supply chain is proposed,and on this basis,a collaborative prediction model of smart tourism supply chain is constructed.Then,taking Jiuzhaigou scenic spot as an example,the Internet search index data was obtained by crawler technology,and the collaborative prediction model was established by combining machine learning model Xgboost and Prophet.The contribution of each smart tourism supply chain member to collaborative prediction was analyzed through the feature importance of each feature in the collaborative prediction model.Finally,the collaborative forecasting model is compared with the single model to explore the prediction performance of collaborative forecasting in smart tourism forecasting.This paper draws the following conclusions: the contribution of smart tourism supply chain members to collaborative forecasting is different.The dominant contribution degree is the tourism destination,followed by the online media marketers,and the last is the tourism transportation.As one of the big data,Internet search data plays an important role in smart tourism prediction.Prophet-Xgboost collaborative prediction model has excellent prediction accuracy,which can effectively help smart tourism scenic spots to better predict the change of tourist flow.CPFR mode can strengthen cooperation among supply chain members,improve operation strategies,reduce unreasonable competition,and empower smart tourism.