Clustering using ordered weighted averaging operator and 2-tuple linguistic model for hotel segmentation: The case of TripAdvisor

•A novel MCDM approach based on OWA and the 2-tuple linguistic model is proposed.•Renamed linguistic quantifiers to reflect customer demand degrees are presented.•Over 50 million TripAdvisor hotel reviews are applied to evaluate its functionality.•The proposed model improves clustering results and linguistic interpretability.•This approach helps to create personalized hotel rankings with customer preferences. With the growth of online tourism, it is important to analyze the reviews left by numerous customers on social networks to improve the hotel's online reputation. Hotel segmentation based on online reviews has attracted an increasing interest from many academics. The problem is that many hotel segmentation models overlook the fact that some customers focus on positive reviews when choosing a hotel, while others focus on negative ones. To address this shortcoming, this paper develops a novel approach to classify hotels using the ordered weighted averaging (OWA) operator, the 2-tuple linguistic model, and K-means clustering. The proposed approach has been evaluated with a real dataset from TripAdvisor, which contains more than 50 million customer online reviews on eight aspects of the hotel. The results show that the proposed model can produce denser and more separated clusters than the model without linguistic quantifiers. From a business point of view, this model enables hotels to distinguish customers' perceptions (from the less demanding to the most demanding) about their eight aspects, allowing them to specify which of them need to be improved and develop strategies more quickly. At the same time, it introduces a new way of ranking hotels online, allowing customers to create personalized rankings of hotels based on their degree of demand for various hotel aspects (better location, cleaner rooms, etc.) rather than the average ratings, so that they can select the most suitable hotels more quickly.