Semantic tools are composed of two main modules: the semantic indexing module and the semantic search module.
- The semantic indexing module provides a service to build a knowledge graph from any dataset selected by the user, provided this data has temporal and geographic properties. This module uses several vocabularies to represent the data in RDF format in a homogeneous way. The data sources to be indexed (or integrated) could be a raster of change (from Thales), of NDVI, or land cover annotation (from DLR). The resulting knowledge graph is stored in a Strabon triplestore. The module is implemented on the platform and accessible through Jupyter-notebook available in the public folder.
- The semantic search module: this module takes into account the kind of data that is looked for. It provides a relevant way to express the user’s interest, the searched area, and the period. The recorded semantic data can be accessed by various means:
- Strabon end-point web interface: https://platform.candela-h2020.eu/semsearch/ep/;
- SPARQL service: Users can make use of semantic tools and libraries to query the store, for instance, Jena (Java framework), SPARQLWrapper, or RDFLib (Python libraries); The query must be sent to https://platform.candela-h2020.eu/semsearch/ep/Query;
- Strabon’s client library.
We provide the user with two additional possibilities to easily perform a semantic search:
- Jupyter notebooks: Notebooks with prepared scripts and semantic queries have been developed and uploaded on the platform.
- Web interface: A web interface has been developed in order to simplify the query that takes into account three dimensions: spatial, temporal, and data-specific.