OpenStreetMap: Quality assessment of Brazil’s collaborative geographic data over ten years Gabriel Franklin Braz de Medeiros 1 , Maristela Holanda 1 , Aleteia Patrícia Favacho de Araújo 1 , Márcio de Carvalho Victorino 2 1 Computer Science Department – University of Brasilia (UnB) Brasilia – DF – Brazil 2 Faculty of Information Science – University of Brasília (UnB) Brasilia – DF – Brazil gabriel.medeiros93@gmail.com, mholanda@unb.br, aleteia@unb.br, mcvictorino@unb.br Abstract. OpenStreetMap is a collaborative mapping tool in which users actively include, transform and exclude geographic data. Consequently, the quality and consistency of the information made available in the tool is of constant concern. To address this issue, this work performs an analysis of some of the quality parameters within OpenStreetMap, with the data referring to the region corresponding to Brazil, over a ten year period, as a source. Analyzing the parameters of Completeness, Logical Consistency and Temporal Accuracy, some basic characteristics of this type of tool can be observed, such as heterogeneity, since mapping does not occur uniformly. 1. Introduction With the advent of Web 2.0, in the early 2000s, Internet users were provided with the capability of creating, changing, and deleting site content in a very dynamic way [Goodchild, 2007]. This event led to the emergence of new techniques and computational methods, which depend on many users for the completion of specific tasks – described as crowdsourcing tools [Tapscott and Williams, 2007]. Some crowdsourcing tasks have customarily been carried out on traditional desktops. However, this method does not always work due to requirements involving the actual physical locations of specific objects. For this reason, a new paradigm called space crowdsourcing has emerged [Zhao and Han, 2016]. Subsequently, the development of smartphone devices with integrated GPS contributed significantly to the emergence of space crowdsourcing, since it allows users to complete tasks according to their physical location. In this context, the OpenStreetMap crowdsourcing tool (OSM), created in 2004 by computer student, Steve Coast, of University College London (UCL), aimed to create a free and editable world map built by volunteers, and released with an open content license [Mark, 2006]. All data from the OpenStreetMap tool can be downloaded for free in vector format, which leads to a widespread use of this data. Given the possible applications for the use of spatial data, such as region mapping, geographic analysis and risk prevention, the issue of information quality is fundamental [Girres and Touya, 2010]. Aside from this paper, few works have analyzed Brazilian OSM data. Thus, this paper performs an Proceedings XVIII GEOINFO, December 04 th to 06 nd , 2017, Salvador, BA, Brazil. p 110-115. 110