MediaGraph

MediaGraph is a content-aware data model and query framework for the representation and retrieval of multimodal knowledge graphs. In this model, the content of the media objects themselves plays a central role in the structure of, and access to, the data — an approach that has been explored in the broader context of modeling multimedia with multimodal knowledge graphs.

Open source

The framework is accompanied by MeGraS, an open-source store for multimodal knowledge graphs that is built on top of the same data model and is demonstrated in the context of the vitrivr multimedia retrieval stack and of multimodal knowledge graph interaction more broadly.

Publications

• MediaGraph: A Content-Aware Data Model and Query Framework for Multimodal Knowledge Graphs — Florian Ruosch, Luca Rossetto — CoRR preprint (2026). doi:10.48550/arXiv.2608.18895
• MeGraS: An Open-Source Store for Multimodal Knowledge Graphs — Luca Rossetto, Florian Ruosch — Proceedings of the 33rd ACM International Conference on Multimedia, MM 2025, Dublin, Ireland (pp. 13644–13647). PDF · doi:10.1145/3747027.3756872
• Beyond the Blob: Demonstrating MeGraS for Multimodal Knowledge Graph Interaction — Luca Rossetto, Florian Ruosch — MMM 2026, Prague, Czech Republic (LNCS 16415, pp. 1614–1620). PDF · doi:10.1007/978-981-95-6963-2_8
• Applications of Multimodal Knowledge Graphs in Modeling Multimedia — Florian Ruosch, Luca Rossetto — MMM 2026, Prague, Czech Republic (LNCS 16412, pp. 636–647). PDF · doi:10.1007/978-981-95-6950-2_45