Extração de dados web como suporte na elaboração de indicadores do turismo de Minas Gerais: uma iniciativa em Big Data

Abstract

The research aims to study the phenomenon called Big Data and the possibility of using web data extraction tools (web scrapers) to help the development of indicators about tourism in Minas Gerais State (Brazil). For that, it was carried out a bibliographical review of authors related to information science to contextualize the subject, as well as to emphasize the role of web information extraction tools. After this step, we used a case study with a web scraper tool to collect data from TripAdvisor, searching for key information about Minas Gerais tourist attractions and turning them into a structured database. Thus, it was possible to analyse information such as the division of tourist attractions by categories from the state and municipalities, the number of evaluations, visitors' profiles, satisfaction levels, and the period of most visits at each of the attractions. To prove the use of the information captured it was carried out a follow-up of the data concerning the Pampulha Architectural Complex with the objective of evaluating a possible impact of its recognition as a world heritage site in the visitors perception. The results showed that it is possible to use data from the platform to monitor actions and create indicators that aim to assist public decision-making. However, there is still need for further discussion about the availability of data delivered by online companies to the final public, which could be used by government agencies. We expect this methodology to assist the state authorities and municipalities to extract strategic information that is already available on the web at low costs, improving actions and ensuring an improvement in the use of public resources in tourism policies