Proc. of the 4th Int. Semantic Intelligence Conf. (ISIC). Lecture Notes in Computer Science. 2025.
Practical business intelligence (BI) over heterogeneous data sources, including relational, vector (e.g., geometries), and raster data (e.g., satellite images), requires interconnecting the data in a semantically coherent manner so that they can be queried and analysed in a uniform way to extract business insights that aid informed decision-making. The Virtual Knowledge Graph (VKG) paradigm addresses the issue of data heterogeneity by relying on an ontology to expose domain knowledge and connecting it via declarative mappings to the underlying data sources. The VKG paradigm has thus far concentrated mainly on relational data. At the same time, only a few works address its combination with vector and raster data, which is especially important within Geospatial Business Intelligence (GeoBI), such as in the context of earth observation (EO), the integration of geographic information systems, and building information modelling (GIS/BIM). However, such a combination presents a considerable challenge for conventional DBs to manage or query efficiently, due to their multidimensional and complex characteristics. In this paper, we address this problem by extending the VKG paradigm to enable interface with specialised array database management systems. We then demonstrate how to utilise BI tools to derive location-based business insights, leveraging both standard semantic technologies and a novel technology that enables a knowledge graph to be accessed via a traditional SQL interface.
@inproceedings{ISIC-2025,
title = "Semantic Enrichment of Location-based Business Intelligence using
Virtual Knowledge Graphs",
year = "2025",
author = "Arka Ghosh and Albulen Pano and Benjamin Cogrel and Diego
Calvanese",
booktitle = "Proc. of the 4th Int. Semantic Intelligence Conf. (ISIC)",
publisher = "Springer",
series = "Lecture Notes in Computer Science",
}
pdf