Building up a Categorical Sentiment Dictionary for Tourism Destination Policy Evaluation

Information technologies and the recent pandemic have reshaped the tourism sector. For tourism policy evaluation that is traditionally measured by evaluation methods, such as questionnaire and interview, we developed a dictionary-based sentiment analysis system on tourism destinations using review data generated by tourists. The system is based on the questionnaire developed by the Korean Ministry of Culture, Sports, and Tourism. Through a series computational method as well as human rating processes, sentiment lexicons, such as ‘sentimentally polarizable’ and ‘categorically discriminable,’ are extracted. The extraction procedures of the sentiment lexicons are presented in the study. On the basis of the dictionary, Haeundae beach, a famous tourism destination in Korea, is analyzed. Our result is encouraging with respect to reshaping the tourism policy evaluation with the text mining method.