Comparative analysis of hotel reviews with text mining methods: The case of Majorca, Antalya, Sharm-el-Sheikh

This thesis focused on the evaluation of tourist reviews of hotels on travel review sites. In this study, it was aimed to measure the characteristics of 3 all-inclusive tourist destinations, users' opinions about the service components of hotels and user's emotions by examining the reviews of the users who visited Antalya, Majorca and Sharm El Sheikh. The study was designed with the content analysis method and data mining techniques were used to assist the content analysis and to process big data. A total of 3588 user reviews, written in English, belonging to the top three hotels in the area of Antalya, Sharm El Sheikh and Majorca, were analyzed. Data mining libraries belonging to R program (such as rvest, tm, syuzhet) were used for data analysis. In the study, the themes and emotions derived from the user comments were significantly similar, but the data of Antalya and Majorca were more correlated. In addition, Sharm El Sheikh had a positive separation from these two destinastions. As a result of this study, it was thought that important data and data processing methods were transferred to tourism enterprises, DMOs and researchers. In order to evaluate the information obtained from the user-generated data for hotels located in Antalya, Majorca and Sharm El Sheikh destinations, methods and information required for low cost evaluation, methods required for quick access to the pattern among the comments made on the travel review sites, the methods and information required to make use of meaningful data obtained from the comments given to the most rated hotels of the regions by the travelers were transferred. Keywords: Text Mining, Tourist Reviews, Content Analysis, All-Inclusive, Sentiment Analysis