Analysis of OpenStreetMap to support an official hybrid database
ISBN 978-85-88783-11-9
Authors
1Borba, R.; 2Strauch, J.; 3Souza, J.; 4Coleman, D.
1UFRJ Email: rogerio.borba17@gmail.com
2IBGE Email: julia.strauch@ibge.gov.br
3UFRJ Email: jano@cos.ufrj.br
4UNB Email: dcoleman@unb.ca
Abstract
Brazil is an emerging country of continental size and disparate conditions. With more than 5,500 municipalities and 27 federal units, the mapping and updating cartographic processing is time consuming, expensive and requires a lot of effort. Within the Government, there are organizations that produce geographic information, such as the Brazilian Institute of Geography and Statistics (IBGE) and the Board of Geographical Service of the Brazilian Army (DSG). However, the lack of technical, financial and human resources is a critical factor that undermines the mapping processes conducted by these institutions. The consequence of this scenario is stagnation of official cartographic production. On the other hand, the use of geographic information produced by initiatives of crowdsourcing in geographical information for public and private institutions ,around the world, is now under way or under study. Indeed, the term Volunteered Geographic Information (VGI) coined by Michael Goodchild turned into a research area and gained attention in academia. Several researches have been and are being conducted in this area. Some research showed that in certain regions, this type of information has a significant variability in relation to their community, production and quality and there are several implicit and explicit factors that influence this variability. For example, Pascal Neis et al. conducted a survey in 12 urban areas in different parts of the world. They found similarities and differences in relation to the database and the user community. Furthermore, their analyses show that socio-economic and income factors may have an impact on the number of active contributors and the data provided in the analyzed areas. There are also works focused on dimensions of quality of VGI database. For example, Haklay conducted research on the positional accuracy of the roads on OpenStreetMap, and compared with the official database of England. He did some considerations and suggestions for future developments and research directions. Jean François Girres and Guillaume Touya investigated the dimensions of the quality of the OpenStreetMap database in France. Michael Goodchild and Linna Li stated that VGI suffers from a general lack of quality assurance. They discussed the issues involved in determining the quality of volunteered geospatial data and described three approaches to support quality assurance. Some researchers investigated the use of volunteered database into official or commercial databases. For instance, David Coleman raised some questions regarding VGI and stated that this mapping initiative is not a solution to all kinds of problems of creating and updating official or commercial geographic database, but it is an important resource that should be taken into consideration. Finally, he argues that research in this area is still at an early level. In Brazil, institutions such as IBGE produce geospatial information for different types of products and applications. Thus require different quality levels, for example, positional accuracy and completeness. The OpenStreetMap is already being used as one of the options in a visualization tool of NSDI in Brazil. However, we still lack a study regarding the integration or use of the OpenStreetMap database into official databases. This work examines how open databases could be incorporated into official databases, thus creating a hybrid database taking into account selection criteria and quality-based filters. This paper also aims to answer the following questions: (1) In which regions these open solutions provide more information? Which regions offer less information? What is the relationship with the user community? (2) What is the quality of that open database? Is it possible to generate a hybrid base? (3) What are requirements to accept the whole or only part of this database? Or discard it? (4) what could be done to create a symbiotic relationship between open and official mapping?
Keywords
VGI; Mappping; Database