관광 빅데이터를 활용한 지역관광 소비지출액 결정요인에 대한 연구

Abstract

This study examined changes in domestic regional tourism occurring after COVID-19, with a particular focus on the Gyeongnam region which has been recognized as a hub for marine leisure and industrial tourism. By doing so, it tried to provide empirical analysis results applicable to the formulation of regional tourism management and marketing strategies by using a non-parametric model. To overcome situations where the linear models commonly used in big data analysis are difficult to apply and to diversify the methodological aspects of big data analysis, this study proposed an additive model to represent nonlinear relationships. Key study procedures are as follows. Firstly, 22 cities in the Gyeongnam region were classified into four clusters based on a cluster analysis of visitor increase and decrease patterns. Subsequently, using an additive model for non-Euclidean variables, the impact of age group-specific visitor ratios and navigation search type ratios, extracted from big data on tourism expenditure by industry, in each of the four clusters were examined. Notable findings include a recent slight increase in the proportion of visitors in their 50s in the Gyeongnam region compared to pre-COVID-19. In Cluster 3, where the most significant differences in tourism patterns pre- and post-COVID-19 were observed, sectors such as accommodation and shopping were significantly related to tourism expenditure. Furthermore, except for one cluster, those aged 50 or older had the most influence on expenditure in most regions. Based on these results, the study suggests that designing models considering the nonlinear characteristics of diverse data can be valuable in big data analysis. Along with this, the study proposes the necessity of formulating big data-based strategies in regional tourism policies and practical fields, taking into account the insights derived from the analysis.