Explainable GeoAI Real Time Data Model for Heterogeneous Datasets: Graph Database Approach

Abstract

Spatial activities are described and linked to the identified place or location. In the age of the Internet of Things (IoT), a vast collection of spatial datasets is emerging. The introduction of GeoAI into spatial data analytics is changing the scope and perspective of analytical capabilities in many ways. Since GeoAI is the merging application of spatial data science, artificial intelligence, and geospatial information science, and is the highest and most advanced application of geoenrichment, intensive heterogeneous data sources have been used. Due to the extensive open data sources generated by mobile devices, sensor data streams from static or moving sensors, satellites, the availability and sharing of data via standard APIs have now increased immensely. In this article, a graph database approach is intensively emphasized to develop an object oriented based explainable GeoAI data model in its various applications. In addition to the available data sources, large amounts of data are currently being generated by various institutions. The issue of sharing and reusing data between institutions is receiving more and more attention for various reasons. Linking datasets between different platforms creates ambiguities for both machine and human. In this article, the research mainly analysis the problems in real-time generated data management of heterogeneous spatial data in the application of GeoAI and provided recommendations.