Sentiment analysis is becoming increasingly crucial for decision-making in the hospitality industry due to the significance of consumer opinions. These opinions, expressing the attitude to services provided, are the data that can be transformed into relevant information to make optimal decisions. However, obtaining customer reviews and conducting sentiment analysis are insufficient to solve management problems. The challenges of operational decision-making motivate the development of review summarization models, which, along with providing a generalized assessment, could also reflect the degree of confidence in it. One of the relevant tools is the Z-number-based approach. Machine learning-based models performed sentiment analysis of hotel reviews. The obtained sentiment scores allowed us to determine aspects of the hotel’s activities (criteria) and the degree of consumer satisfaction. An integrated assessment of each hotel’s activities (criteria) – Z-evaluations are obtained. After calculating the criteria weights based on the Z-preference matrix reflecting the relative importance of each criterion and application of the Z-SAW method, each hotel’s activities were assessed to determine the best one. This approach, which integrates sentiment analysis and calculations with Z-numbers, can be applied in the hospitality industry but also in other areas of the service sector.