In our master thesis we developed an approach for analyzing the choice of tourist destinations based on online tourist reviews. In our work, we studied the influence of earlier visits on later visits and, in relation to this, the influence of the attribute location details. In studying the influence of the location details attribute, we wanted to find out how the types of attractions of previous visits influence the selection of Slovenian tourist attractions and municipalities. We tackled the design of the general approach based on data from the TripAdvisor platform, using market basket analysis methods and with the use of co-occurrence matrices. Performance was evaluated with ROC and PR curves and with the calculation of the lift value. The approach, which has proven to be the most effective and is proposed in our master thesis, includes several analyzes that provide analysts with an overview of different profiles of tourists and enable them to market tourist destinations more efficiently. The developed approach can be generalized to the data of any country, and its application makes it possible to improve tourism management and thus increase the number of visits to tourist destinations and consequently increase profits.
|
