Abstract

The aim of this study is to propose an exploratory methodology for automatically estimating multi-dimensional images of travel destinations based on unsupervised learning using a large-scale travel blog corpus so that it can be used for longitudinal and cross-sectional studies. First literature reviews on the construal level theory conducted to clarify the relationship between the travel experiencer's level of travel destination interpretation and potential travelers' travel destination selection decision-making and instead of surveys, the value of travel blogs as message information on the travel experiencer's travel destination interpretation are useful data resource for automatically estimating multi-dimensional images was discussed. Next, by applying web crawling, natural language processing technology and the unsupervised learning-based machine learning algorithm principal component analysis, the frequency of co-occurrence of the travel destination named entity-subjective adjective pair is vectorized to estimate the multi-dimensional images of travel destinations, and visualize the image property structure of individual travel destinations through network centrality analysis by referring to the destination image attributes of Echtner and Ritchier (2003). Finally, the validity of this methodology was demonstrated by applying to the case of Changwon. The results of this study can contribute to gaining insight in making demand-oriented regional tourism vitalization policy decision-making and can be used for large-scale travel blogs corpus-based convergence follow-up studies.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call