Medindo a acessibilidade: uma perspectiva de Big Data sobre os tempos de espera e tarifas do serviço da Uber

• Ano de defesa: 2020
• Autor(a) principal: Insardi, André
• Orientador(a): Strehlau, Suzane
• Banca de defesa: Ponchio, Mateus Canniatti, Francisco, Eduardo de Rezende
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Escola Superior de Propaganda e Marketing
• Programa de Pós-Graduação: Programa de Mestrado Profissional em Comportamento do Consumidor
• Departamento: ESPM::Pós-Graduação Stricto Sensu
• País: Brasil
• Palavras-chave em Português: comportamento do consumidor; Big Data; acessibilidade; mobilidade urbana; estatística espacial
• Palavras-chave em Inglês: consumer behavior; Big Data; accessibility; urban mobility; spatial statistics
• Área do conhecimento CNPq: CIENCIAS SOCIAIS APLICADAS::ADMINISTRACAO
• Link de acesso: http://tede2.espm.br/handle/tede/507

In a world transformed by the sharing economy and rapid technological evolution, one of the most affected markets is urban mobility, crowdsourcing transportation companies like Uber, 99, Lifty and others have changed the way people get around, what comes impacting the auto industry and the cities where the service is offered. In addition to these impacts, this new way of getting around produces a new study artifact, the digital trail of displacement, which can be understood from the perspective of Big Data, for example, when getting around the city through a shared ride service, the user and driver adjust the fare via the application, installed on their smartphones, where all the digital trace of the service provided is collected, such as origin, destination, route and other data. A theme that has been of interest to geographers, urban planners and sociologists over the years is that of accessibility, a measure that reflects the spatial development that consists of the transport network and the distribution of opportunities, reflected by the uses and occupation of urban land. This measure is also used by companies in different segments to plan investments and support decision making, such as some retail segments that use the distribution of public transport to plan the positioning of their stores. Therefore, this study seeks to relate the approach of Big Data with the concept of accessibility from the new data produced in the provision of services by transportation crowdsourcing companies. An exploratory study of the correlation of the tariff and the waiting time for the service and dynamic tariff is proposed, with socioeconomic variables from the city of São Paulo with the intention of exploring the use of these measures as an accessibility proxy. For this, the study proposes the creation of a database with the average of the estimated tariffs and the waiting time of the service, added to a set of socioeconomic variables and transport infrastructure. From this base it is proposed to develop MLR linear regression models, using the stepwise method selecting the most significant variables in the model and checking if there is a spatial pattern of the variables through the Moran I test, ending with a SAR autoregressive spatial model. All the models developed showed variables with a high degree of significance, such as: percentage of non-whites, amount of bushes and population density, in addition to Rsquare in the range of 0,80.