A study on the effect of imbalanced data in tourism recommendation models

Abstract

Purpose This paper aims to study the effect of imbalanced data in tourism quality models. It is demonstrated that this imbalance strongly affects the accuracy of tourism prediction models for hotel recommendation. Design/methodology/approach A questionnaire was used to survey 83,740 clients from hotels between five and two or less stars using a binary logistic model. The data correspond to a sample of 87 hotels from all around the world (120 countries from America, Africa, Asia, Europe and Australia). Findings The results of the study suggest that the imbalance in the data affects the prediction accuracy of the models used, especially to the prediction provided by unsatisfied clients, tending to consider them as satisfied customers. Practical implications In this sense, special attention should be given to unsatisfied clients or, at least, some safeguards to prevent the effect of the imbalance of data should be included in the models. Social implications In the tourism industry, the strong imbalance between satisfied and unsatisfied customers produces misleading prediction results. This fact could have effects on the quality policy of hoteliers. Originality/value In this work, focusing on tourism data, it is shown that this imbalance strongly affects the prediction accuracy of the models used, especially to the prediction of the recommendation provided by unsatisfied customers, tending to consider them as satisfied customers; a methodological approach based on the balance of the data set used to build the models is proposed to improve the accuracy of the prediction for unsatisfied customers provided by traditional services quality models.

Open Innovation through Customer Satisfaction: A Logit Model to Explain Customer Recommendations in the Hotel Sector

“…. + βm Xm, where m is the number of independent variables X1, X2, X3, and X4, which in this study are: customer service, facilities, cleanliness, and price-quality ratio [66].…”

“…For these reasons, a possible study in the future should include more attributes in order to know what the hotel recommendation variables are and how the cleanliness measures could change the importance of the model's quality-price ratio. Furthermore, these studies in the current pandemic context should include balanced samples with respect to the recommendation variable [66].…”

An empirical study on the imbalance phenomenon of data from recommendation questionnaires in the tourism sector

“…Therefore, it is not known how the imbalance data affects the prediction models and this generates inaccuracies in the results in the accommodation field. While it is true that there are some studies that identify the existence of imbalanced data in hotels (Chawla et al, 2002(Chawla et al, , 2004Li and Sun, 2012;Fern andez-Muñoz et al, 2019) none of them go as far as applying it to predictive models.…”

“…The existence of imbalance data is related to a quantitative analysis of the sample or data collected, and its treatment or not will produce competitive advantage differences among the companies and their marketing campaigns. Quality hotels studies assume there is an equilibrium between the number of satisfied customers (who would recommend the hotel) and unsatisfied customers (who would not recommend the hotel), but this is usually far from true (Fern andez-Muñoz et al, 2019). The existence of imbalanced samples in which the number of satisfied customers is larger than the number of unsatisfied customers directly affects the predictive capacity of the models and has implications for hotel management, development of marketing campaigns, construction of a brand image and the development of competitive advantage.…”

“…In this regard, some previous studies present isolated solutions to solve this, but it is not assumed as an endemic situation within data coming from recommendation surveys (Fern andez-Muñoz et al, 2019;Li and Sun, 2012).…”