| 研究実績の概要 |
Spatial data processing and analysis have several applications in different domains including marketing, disaster management, autonomous driving, robotics, etc. In many cases, spatial data is huge and require real-time processing to avoid heavy financial and human loss. In this project, we researched and developed a scalable and real-time data stream management system GeoFlink for the efficient processing of spatial data streams.
In the financial year 2020, we developed a preliminary GeoFlink framework and wrote a conference paper which was published in a top ranking conference CIKM2020. GeoFlink extends Apache Flink to support spatial data types, indexes, and continuous queries over spatial data streams. To enable efficient processing of spatial continuous queries and for the effective data distribution across Flink cluster nodes, a gird-based index is introduced.
In the financial year 2021, the design and implementation of GeoFlink framework was completed and we wrote a Journal paper which was published in IEEE Access. In this year, GeoFlink was extended to support different spatial objects besides spatial Point, i.e., Polygon and LineString. Additionaly, extensive experiments were performed to evaluate the GeoFlink framework. The complete framework is available on Github as an open source and is registered as an AIST intellectual property. GeoFlink supports spatial range, spatial kNN and spatial join queries on spatial geometries (point, linestring and polygon).
|
