Studying the effectiveness of investments in hotel services using customer sentiment analysis

Abstract

The main aim of the article is to assess investments in hotel services in Poland in 2018–2020. The assessment was carried out by examining the correlation between financial outlays in the hotel industry in powiats (counties) and voivodeships of Poland and customer opinions expressed in internet entries on the websites of individual tourist facilities. The methodology for testing the sentiment of text documents is an unsupervised classification algorithm applied to the Polish language. The classification task consists of the unsupervised assignment of a text document to one of three sentiment classes: positive, neutral, or negative. The research is based on opinions collected from 906 hotels and the amount of financial investment in individual hotel segments (categories) in the surveyed years on a powiat basis. Financial investment data were gathered from the official statistical yearbook websites and internet entries were scraped from the websites of individual tourist facilities. The analysis shows a positive relationship between the size of investment in hotel services and the sentiment of customers' opinions. The scientific contribution of the research is threefold. Firstly, an attempt is made to assess the efficiency of investments in hotel services in Poland in recent years. Secondly, the approach applied omits the downsides of other popular models of assessing the quality of services. Thirdly, the algorithm used quite precisely finds the categories required for a different kind or broader analysis of the quality of hotel services.