Determination of Customer Satisfaction by Text Mining: Case of Cappadocia Hotels

Özet

Purpose – The aim of this study is to analyze the comments of travel web sites on 28 hotels that have 4 or 5 stars, operating in Cappadocia region by text mining and to determine the satisfaction level of customers from hotels. For this purpose, text mining methods were used and 10 263 comments were downloaded and analyzed. Design/methodology/approach – The research focuses on 5 and 4-star hotel reviews of the Cappadocia region in the English language version. For this study, 10263 comments were obtained from 28 different hotels. The excel file created in the context of data acquisition was added to the RapidMiner data mining program. Then, a model was created for the individual separation of the words mentioned in the headlines and comments. The aim is to make the word frequencies by determining the words in the comments as attributes, and also to make them suitable for the association analysis. Findings – It was determined that the tourists made comments on the room experience, staff and hotel dimensions. They have stated that the rooms were great, clean, gorgeous, amazing, unique and spacious. The hotel staff was found to be helpful and friendly. It is concluded that hotels are generally clean, beautiful and unique. It can be said that there is a high level of satisfaction from the hotels. Discussion – The fact that the data obtained by the method of text mining has facilitated generalizations. This method can be defined as the most common factors that tourists perceive when expressing their opinions and sharing their experiences online. In this respect, they have directly received the service quality, physical condition, etc., they expressed their comments about many factors as easily as they felt. It is thought that providing online environments with such facilities is important in obtaining reliable and valid information.