http://lattes.cnpq.br/5456214966384238; ACIOLI FILHO, José Amilton Moura.
Resumo:
With the evolution of the web and information systems, organizations have obtained data of various formats, structures and types, specially the spatial one. Due to different characteristics presented in data, such data have been stored in heterogeneous data sources. Therefore, it is needed to increasingly invest in solutions that can integrate and analyze these data from different sources. Some of these solutions can analyze the spatial component of data; however, this analysis of spatial data is limited either by the data type or spatial functions supported. In this work, the problem of spatial data integration from heterogeneous data sources is addressed, either with or without using schemas, using SQL language. This is an open issue in the area of spatial data integration, since existing solutions present many limitations, such as the query language used, the ways to access data, the technologies that can be integrated, the available functions set and the spatial data types supported. Aiming at solving this problem, the GeoDrill solution was developed, which is an extension of the Apache Drill that supports all standard spatial functions provided by the OGC (Open Geospatial Consortium) through the SQL language. The GeoDrill can perform queries on data with or without schema. In order to validate the capacity of GeoDrill to integrate data, an experiment was conducted to analyze its functionalities and performance. The obtained results indicate the GeoDrill solution is able to integrate spatial data from heterogeneous data sources. Hence, it appears to be a suitable alternative for solving part of the existing limitations in this research field.