Uma solução de extração e georreferenciamento de anúncios imobiliários da internet

Abstract

The growth of urban territory is something that cannot be controlled, becoming a difficulty for municipal organizations. With the usage of geographic information systems and urban simulation, a new way to make decisions in city planning emerges. Although there is a big gap for it happen naturally by municipalities, this is due to the lack of georeferenced databases for the study of the urban network. As the Internet is a place full of information, it becomes a great source to extract the desired content. From the study of real estate sales, it is possible to identify changes in the city by modifying the sale price and expanding the territory by the emergence of new real estate in a given region. This work proposes to create a solution that enables the extraction of information from real estate sites, such as the address and sale value and turn the address into geographic coordinates, latitude, and longitude, providing a database to be used in studies of the city. In this work it was made use of real estate sites in the city of Ponta Grossa - Paraná to extract unstructured information, addresses, and sales values so that this content could be turned into information applicable to urban simulation systems and geographic information. To transform address into geographic coordinates it became necessary to use geolocation services such as Google Maps and Open Street Map. As this information does not follow a pattern the document-oriented format was chosen so it could store the found data. The application proved to be efficient due to the number of ads extracted and the user's independence to extract and georeferenced the information.