Abstract
As a consequence of the exponential growth in online data, tourism sector has experimented a radical transformation. From this large amount of information, opinion makers can be benefited for decision making in their purchase process. However, it can also harm them according to the information they consult. In fact, being benefited or harmed by the information translates into greater or lesser satisfaction after the purchase. This will largely depend on the published opinions that they take into account, which in turn depend on the value of the opinioner who publishes said information. In this paper, the authors propose a methodology that integrates multiple decision-making techniques and with which it is intended to obtain a ranking of hotels through the opinions of their past clients. To do this, the customer value is obtained using the Recency, Frequency, Helpfulness model. The information about the users found in the social networks is managed and aggregated using the fuzzy linguistic approach 2-tuples multi-granular. In addition, we have verified the functionality of this methodology by presenting a business case by applying it on TripAdvisor data.
Highlights
Due to the expansion of new technologies and the intensive use of social networks, clients have increased the publication of their opinions in forums and social networks to value products and services, this has strengthened the relationship with the client through the Internet [1]
The electronic word of mouth in this type of customer reviews has a great impact on the process of purchase or selection of services by the consumer [2,3,4,5]
Throughout this work we have raised the idea that a ranking of a set of hotels can be obtained based on customer opinions, either through questionnaires or reviews
Summary
Due to the expansion of new technologies and the intensive use of social networks, clients have increased the publication of their opinions in forums and social networks to value products and services, this has strengthened the relationship with the client through the Internet [1]. The customer has access to an infinity of other users opinions who have enjoyed of the products and services in which they are interested For that reason, it is worth considering the customer’s value based on their online opinions and how the rest of the users notice them as useful or reliable. To obtain an overall assessment of the product and/or its characteristics, which normally represents the degree of agreement or disagreement with each characteristic through Likert scales [9, 10]. On a 5-point scale, responses are usually ‘‘strongly disagree’’, ‘‘disagree’’, ‘‘neutral’’, ‘‘agree’’ and "strongly agree" These responses are characterized by the uncertainty and blurriness of the perception they represent [13], since the same concept can indicate very different perceptions [14]. Some authors [15] consider that an approach based on the use of linguistic evaluations would be better to model these human perceptions than that of conventional numbers (crisp)
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
