Language-Segmented Study Based on TripAdvisor Reviews Related to Memorable Tourist Experiences

摘要

The Memorable Tourist Experience(MTE)is a scientific concept within the studies on tourism that is developed based on several related constructions:Perceived Confidence,Sincerity,Authenticity,and Satisfaction.This work takes this model established by the work of Dr.Babak Taheri in 2018 on Monuments World Heritage of UNESCO,adopting an alternative data collection method to the face-to-face survey.Therefore,this work takes as a source of data the reviews collected in the recommendation platform TripAdvisor,working the same constructions of the MTE,with the collection of similar terms and the relationships between them.In order to highlight the terms,a first step is established with the use of Natural Language Processing(NLP),followed by the use of Machine Learning(ML)techniques to generate the relationships between the constructors defined in the models.The study makes a comparison using the method,in immaterial nature such as a flamenco show in the city of Seville;Flamenco has been declared by UNESCO an intangible World Heritage Site since 2010.The results of the study go in two directions:on the one hand to find similarities in the study of the specific MTE of both monuments with the hypotheses worked in the original model of Taheri.In addition to highlighting possible distinctive elements of each case and,and furthermore within the value contribution of the visit when it is led by an official tour guide,on the other hand,give presence to the model of obtaining data by reviews as a complementary data source of any tourist study.The data collection and analysis from both NLP and ML techniques permit the scientific study and the tourist operators to develop better value propositions to users and understanding of heterogeneous behaviors in the tourism industry.The study of reviews within the MTE allows identifying the stimulus that leads the user to choose an activity and hire it.These studies are extendable to other industries and business models,given the importance that references acquire within the consumer willing to buy.For the scientific community,the use of ML is a solid way to initiate studies on behavioral models,supplement them,and accept or reject hypotheses.When the source of the data is taken from free expressions,such as reviews,the appearance of bias in the behavior is attenuated.