Fabre, Renaud

Preferred name
Fabre, Renaud

Settings

Now showing 1 - 1 of 1
  • Thumbnail Image
    Some of the metrics are blocked by your 
    consent settings
    Item type:Publication,
    Knowledge Graphs – The Future of Integration in CRIS Systems for Uses of Assistance to Scientific Reasoning
    (euroCRIS, 2024-05-16)
    ;
    Knowledge graphs (KG) are increasingly coming into focus as they provide a powerful method for data integration and knowledge representation. Their semantic data model, which represents knowledge in terms of entities, attributes, and relationships between those entities, applies well to descriptive encyclopedic uses, but encounters challenging limitations in scientific knowledge applications where support for contested knowledge categorizations in research is poorly applied. In the context of the platformization of science, especially in the context of CRIS systems, we have observed that knowledge graphs can be used to combine the diverse data sources and data formats that exist in the research landscape and create unified and connected data models. It therefore enables researchers, administrators and other stakeholders to access comprehensive and consistent information relevant to their work. The following paper examines the specific role of KG in the future of data integration in CRIS systems in supporting scientific thinking. It highlights the advantages and disadvantages of the current features of KG. Advantages and disadvantages include flexible knowledge modeling, support for semantic queries, and interoperability with other data sources and formats. Systemic limitations consist mainly in the methodological and technological expression of controversies and scientific disagreements, which significantly limits the potential of the scientific classical investigation of relatedness, identity and categorization of controversial new ideas using knowledge graphs: this represents a serious limitation to the use of the KG of CRIS. The paper presents advances in solutions to support scientific thinking, with various use cases and best practices for implementing KG in CRIS systems, enabling research institutions and scientific organizations to improve their data analysis and support of scientific thinking. Concluding remarks concern ongoing work and new results; finally, we discuss the pace of challenges that open up new approaches to supporting scientific thinking that are currently opening up the interaction of large language models and the KG in new technologies.