Study of Google Popularity Times Series for Commercial Establishments of Curitiba and Chicago

• Yuri C. B. Neves UTFPR
• Mozart P. Sindeaux UTFPR
• William Souza UTFPR
• Nádia P. Kozievitch UTFPR
• Antonio A. F. Loureiro UFMG
• Thiago H. Silva UTFPR

Resumo

Urban computing is a recent area of study that helps us to understand the nature of urban phenomena. In this sense, an important aspect to study is the dynamics of commercial establishments popularity in the city. Recently, Google launched a new service that provides popularity time series of some commercial establishments in several cities. This is a valuable source of data that allow us to better understand the dynamics of establishments popularity, helping to change our perceived physical limits about the city, which can enable the development of new applications and urban services. The results of this study are: (1) characterization of Google popularity time series for bars and restaurants in the cities of Curitiba/Brazil and Chicago/USA. Among the results, we find cultural characteristics of these cities, as well as a favorable clustering of similar venues based on the temporal pattern of popularity; (2) evaluation of reproduction of Google popularity time series using Foursquare data. In this evaluation, we found evidence that Foursquare data might be used for this purpose. This means that for places where Google does not offer this service data from Foursquare, or other source, could be used. This enables the exploration of a greater number of establishments in, for example, a new venue recommendation engine.

Palavras-chave: Computação Urbana, Google, Séries Temporais, Popularidade de Locais, Caracterização, Foursquare, Redes Sociais