News du W3C en 10 minutes par Pierre-Antoine CHAMPIN du W3C / INRIA / Université de Lyon
- Nouvelle recommandation RDF-canonicalization: https://www.w3.org/TR/rdf-canon/
- Outil pour tester la canonicalisation en ligne: https://champin.net/2023/sowasm/?oformat=application%2Fcanonical-n-quads
- Recommandation proposée DCAT 3: https://www.w3.org/TR/vocab-dcat-3/
- Le Community Group "Dataspaces": https://www.w3.org/community/dataspaces/
- Un document d'introduction sur Verifiable Credentials: https://www.w3.org/TR/vc-overview/
Explainable Temporal Fact Validation Through Temporal Constraints Discovery in Knowledge Graphs par Fatiha SAIS du LISN / Université Paris Saclay
Résumé : To face the challenge of validating or refuting information, Knowledge Graphs (KGs) are a means for verifying them from statements in the KG. A feature that has yet to receive attention is the temporal validation of these statements (i.e. verifying whether a statement is true in a given time interval). I will present our work that introduces a new explainable approach that leverages the power of KGs to classify user-inputted facts by assessing their temporal validity or refutation. Our approach is based on the discovery of temporal constraints that express the temporal consistency of fact along with the entity timeline. Allen’s relations between temporal intervals and extends these relations to time sequences. We evaluated our symbolic framework on one of the largest publicly available KGs and compare its performance across multiple hyper-parameters with the neuro-symbolic extension that we propose.
Articles associés:
- Thibaut Soulard, Joe Raad and Fatiha Saïs. "Validation Temporelle Explicable de Faits par la Découverte de Contraintes Temporelles Complexes dans les Graphes de Connaissances." 35es Journées francophones d'Ingénierie des Connaissances (IC 2024) @ Plate-Forme Intelligence Artificielle (PFIA 2024), Jul 2024, La Rochelle, France.
- Fatiha Saïs, Joana E Gonzales Malaverri, Gianluca Quercini. MOMENT: temporal meta-fact generation and propagation in knowledge graphs. Proceedings of the 35th Annual ACM Symposium on Applied Computing, Mar 2020, Brno, Czech Republic. ⟨10.1145/3341105.3374021⟩. ⟨hal-04420928⟩