Data Management and Knowledge-Driven AI
@dmki.bsky.social
32 followers 82 following 6 posts
DMKI is a research lab at TU Wien with the ambition to bring meaning to large amounts of data. https://dmki-tuwien.github.io/
Posts Media Videos Starter Packs
dmki.bsky.social
Maxime Jakubowski presented the paper "Bridging Gaps in RDF Validation – Insights and Innovation Opportunities in RDF Validation Practices" at SEMANTiCS 2025 (@semantics-conf.bsky.social) in Vienna, Austria.

Authors: M. Jakubowski, D. Tomaszuk, K. Hose

Paper: ebooks.iospress.nl/doi/10.3233/...
Reposted by Data Management and Knowledge-Driven AI
katjahose.bsky.social
Honored to have received a Distinguished Associate Editor Award at @vldb.bsky.social 2025!

It was a pleasure to contribute to the review process, and I'm truly grateful for the recognition.

Congratulations to all awardees!

#VLDB2025
Reposted by Data Management and Knowledge-Driven AI
pvldb.bsky.social
Vol:18 No:12 → Smart SPARQL Advisor: Guiding Users in Query Formulation with Performance Prediction
👥 Authors: Abiram Mohanaraj, Matteo Lissandrini, Katja Hose
📄 PDF: https://www.vldb.org/pvldb/vol18/p5295-mohanaraj.pdf
Thumbnail: Smart SPARQL Advisor: Guiding Users in Query Formulation with Performance Prediction
dmki.bsky.social
The work written by Johannes Schrott, Rainer Meindl, Christian Lettner, Stefan Hammer, and Magdalena Leitner provides a novel approach for the knowledge graph-based definition of parameterized data quality metrics and can be downloaded at www.vldb.org/2025/Worksho....
www.vldb.org
dmki.bsky.social
Yesterday, our colleague Johannes Schrott presented the paper "Dynamic Knowledge Graph-based Measurement of Data Quality" at the 14th QDB Workshop held in conjunction with VLDB 2025.
Reposted by Data Management and Knowledge-Driven AI
pvldb.bsky.social
Vol:18 No:6 → PlanRGCN: Predicting SPARQL Query Performance
👥 Authors: Abiram Mohanaraj, Matteo Lissandrini, Katja Hose
📄 PDF: https://www.vldb.org/pvldb/vol18/p1621-mohanaraj.pdf
Thumbnail: PlanRGCN: Predicting SPARQL Query Performance
Reposted by Data Management and Knowledge-Driven AI
pvldb.bsky.social
Vol:18 No:8 → The Limits of Graph Samplers for Training Inductive Recommender Systems
👥 Authors: Theis Jendal, Matteo Lissandrini, Peter Dolog, Katja Hose
📄 PDF: https://www.vldb.org/pvldb/vol18/p2496-jendal.pdf
Thumbnail: The Limits of Graph Samplers for Training Inductive Recommender Systems
dmki.bsky.social
Our team members Koumudi Ganepola and Johannes Schrott attended this year's VLDB Summer School at the University of Wrocław in Poland. Besides lectures from top researchers, they had the opportunity to present their current work in the form of posters.
Impressions from a week full of experiences:
Reposted by Data Management and Knowledge-Driven AI
katjahose.bsky.social
Today, I had the great pleasure of giving an introductory lecture on knowledge graphs at the #BILAI Summer School 2025 in Klagenfurt. Great energy and engagement!
#knowledgeGraphs #RDF #SHACL #SPARQL #LLM

www.bilateral-ai.net/doctoral-sch...
Reposted by Data Management and Knowledge-Driven AI
katjahose.bsky.social
Honored to receive the IFIP TC2 Manfred Paul Award (2024) with Kashif Rabbani and @kuzeko.bsky.social for our VLDB 2022 paper on generating #SHACL shapes from large knowledge graphs. Scales to billions of triples, mining meaningful patterns using support and confidence.
www.vldb.org/pvldb/vol16/...
dmki.bsky.social
Maxime Jakubowski presented our work "Towards Generating Synthetic EHR Knowledge Graphs — a Probabilistic Approach" today, at the 1st GOBLIN Workshop on Knowledge Graph Technologies in Leipzig, Germany.

www.dbpedia.org/events/gobli...

/cc @maximejakubowski.bsky.social

#EHR #RDF #FHIR
Reposted by Data Management and Knowledge-Driven AI
matrig.net
📣 Join us at IBM Research for a PhD position on Reliable Multimodal LLMs, combining language, vision & advanced reasoning!

Located in Zurich🇨🇭 & co-supervised by @katjahose.bsky.social at TU Vienna🇦🇹 as part of the cool new MSCA ARMADA Doctoral Network

Apply at www.zurich.ibm.com/careers/2025...
🧪
Banner of the ARMADA website:
"ARMADA
Reliable Conversational Domain-specific
Data Exploration and Analysis"