Search

According to the world economic forum report, around 70% of generated data are not used. This limitation of usage is mainly due to the lack of interoperability and linking of data that resides in isolated silos. Indeed, the overwhelming amount of data has worsened heterogeneity problems, as have the types of sources generating data in heterogeneous formats and different semantics. Those data related problematics are frequent in the domain of Earth Observation (EO). Earth observed data use different terminologies, which are difficult to reconcile because they reflect overlapped disciplines. These issues lead to misunderstandings and inefficient exchange and management of data in terms of access, pricing, and data rights, which can hamper environmental phenomena understanding. Virtual Knowledge Graph (VKG), allows semantic integration of existing data sources into a wide Knowledge Graph. In this work we propose a knowledge hypergraph-based approach for data integration and querying, with an application to Earth Observation data. Our proposal takes place in two phases (1) a knowledge hypergraph-based virtual data integration and (2) a hypergraph-based query processing. The first phase allows to generate a virtual knowledge hypergraph consisting of RML mappings between an ontology and the data. The second phase consists of enhancing the user’s query by extracting and consolidating a global view of data from different sources based on the generated knowledge hypergraph. The proposed approach is implemented in the Onto-KIT tool (Ontology-based Knowledge hypergraph data Integration and querying Tool) and evaluated through real use case studies. The obtained results show that our proposal enhances query processing in terms of accuracy, completeness, and semantic richness of response.

In various domains, the continually increasing volume of heterogeneous data generated from different sources is overwhelming. However, the exploitation of this data is still limited while, in most cases, sources are not interoperable, and hence, data are not linked. These issues are due to the semantic, syntactic, and schematic heterogeneity of the data. In this work, we propose a semantic virtual Data Integration TOol based on Semantic Enhancement and RML (RDF Mapping Language) mappings called DISERTO. It automatically generates RML mappings through domain ontology and thesaurus to categorize the semantics behind the data. It follows three main steps: (1) extracting relevant metadata (data schema), (2) mapping metadata to a domain ontology by taking advantages of RDF quads, and finally (3) generating RML mappings. To validate the tool, we provide a case study based on real data provided by the Sahara and Sahel Observatory (OSS) and the National Oceanic and Atmospheric Administration (NOAA). It shows a semantic annotation followed by an RML mapping generation of a raster input image and CSV files.

Earth observation (EO) systems play a significant role in environmental monitoring and the prediction of natural disasters. These systems generate a massive amount of heterogeneous data stored in different formats. The exploitation of this data is still limited while, in most cases, data are not linked, and sources are not interoperable. Hence, data cannot be exploited as an interoperable global knowledge graph to have more in-depth analyzes of environmental phenomena. Ontology, as a knowledge representation formalism, is a promising solution for the semantic interoperability between this data. In this work, we present a modular ontology for environmental monitoring developed based on an original agile methodology. The so-called MEMOn (Modular Environmental Monitoring Ontology) aims to support semantic interoperability, data integration, and linking of heterogeneous data collected through a variety of observation techniques and systems. We also present real use case studies to show the usefulness of the proposed ontology.

International audience; The increasing volume of data generated by earth observation programs such as Copernicus, NOAA, and NASA Earth Data, is overwhelming. Although these programs are very costly, data usage remains limited due to lack of interoperability and data linking. In fact, multi-source and heterogeneous data exploitation could be significantly improved in different domains especially in the natural disaster prediction one. To deal with this issue, we introduce the PREDICAT project that aims at providing a semantic service-oriented platform to PREDIct natural CATastrophes. The PREDICAT platform considers (1) data access based on web service technology; (2) ontology-based interoperability for the environmental monitoring domain; (3) data integration and linking via big data techniques; (4) a prediction approach based on semantic machine learning mechanisms. The focus in this paper is to provide an overview of the PREDICAT platform architecture. A scenario explaining the operation of the platform is presented based on data provided by our collaborators, including the international intergovernmental Sahara and Sahel Observatory (OSS).