Tourist segment and attraction characterization through data mining and network analysis on TripAdvisor

This study aims to understand tourist segments and attractions through a replicable methodology for other Brazilian states, expanding the understanding of visitors' perception and facilitating decision- making by planners and managers of tourism destinations, as well as supporting the development of public policies for the sector. The methodology is developed through the application of textual data mining techniques, statistical textual analysis and network analysis in a qualitative database evaluating visitors left on TripAdvisor. Information resulting from comments from social media specialized in tourism is used in order to analyze visitors' discourse about Piaui's tourist attractions. These were achieved through the use of Big Data applied to the traces left by travelers on TripAdvisor, regarding tourist attractions. The qualitative database was treated usingfree software aimed attextual analysis, data processing and network analysis. It is concluded that, based on the results, it is possible to understand tourism demand in Piaui and propose a discourse analysis methodology associated with spatial analysis capable of substantiating the management of destinations, attractions and the development of public policies.