ASSESSING TOURISM AND REGIONAL ECONOMIC INEQUALITY IN CHINESE CITIES: SPATIAL ANALYSIS BASED ON POI BIG DATA AND MACHINE LEARNING

Abstract

This paper presents an overall examination of the interlinkage relationships concerning regional policy, tourism, and economic inequality through the Chinese case. The research objectives are to evaluate the causal effect of the Western Development Strategy (WDS) and provide information regarding the spatial distribution of tourism facilities at the level of Beijing city through the application of contemporary data and advanced methodologies. To achieve the objectives of this study, there are two interconnected empirical research components. The first research component provides a compilation of findings from existing spatial Regression Discontinuity Design (RDD) studies concerning the impact of the Western Development Strategy (WDS). In addition, descriptive statistics from the years 2000 to 2020 are used. The findings of the mentioned researches confirm the existence of a positive effect of the WDS regarding the tourism sector across the targeted regions. This effect is demonstrated through the relative enhancement of the gross regional product’s tourism revenue part at a level of approximately 5.9%–6.7 %. The results of the mechanism approach confirm the indirect support of the WDS regarding the tourism sector through the enhancement of the relevant investments in the sector’s infrastructure and the extension of the Tax Incentive Schemes. The second research component investigates the spatial distribution of tourism and leisure facilities in the primary city districts of the city of Beijing through the application of POI big data information concerning the relevant sector. Additionally, the findings of the research will be used concerning the information of the relevant sector’s POI big data. Machine learning algorithms and decision trees will be employed for the identification of the best locations suitable for the allocation of tourism facilities. The accuracy level of the model achieves the remarkable figure of approximately 83.5 %. The four basic factors that affect the spatial distribution of tourism facilities are hotel density, vicinity to shopping malls, transport accessibility levels, and the relevant sector’s POI big data information regarding the city’s relative population. The results can contribute to the development of an empirical standard regarding the RDD method in the field of tourism economics and the application of POI big data information concerning AI through the enhancement of the effect of regional policy concerning the mitigation of regional inequality levels.