This effort aims to develop a knowledge graph that integrates physical parameters and geologic data from bodies across our solar system.
Before installing and executing the files, a platform for hosting the knowledge graph must be established. For the purposes of this project, GraphDB has been selected. Python interacts with GraphDB via its REST API. With appropriate modifications to the request formatting, alternative graph storage solutions may also be suitable.
After the istallation the first that needs to be done is to load the solar system ontology. For the scope of this project Protege tool was used for the design of the ontology.
Ontology file in turtle format can be found here.
Executing populate_main.py initiates the population of the knowledge graph, beginning with physical parameters followed by geocontacts and geounits. While the insertion of physical parameters is completed almost instantaneously, loading the GeoJSON files can be time-consuming. For reference, on a typical home setup, this process may take approximately 17 minutes. Once the knowledge graph has been populated, SPARQL and GeoSPARQL can be used to query and extract semantic and spatial information from the dataset.
Physical parameters data and documentation be found here.
Geologic data and documentation can be found here.
Discovering the geologic relations among all features of an astronomical object can be a tedious an time consuming task. For that purpose two tools developed by the University of Athens were used to speed up the process:
DRG repository is available here.
JedAI-spatial repository is available here.
Running get_geometries.py generates an N-Triples file for each astronomical object. These files are then can be processed with the DGR tool, which performs geometry deduplication, validation, and formatting in preparation for input to JedAI-spatial. Finally, JedAI-spatial can be used to automatically identify geologic relationships among the features of each astromical object.
The resulting n_triples files with all the relationships can be found here.
Below we can see the geologic features of Io and Moon plotted.
