Applied and Interpretable
Machine Learning

Research

Structured and scientific data

Much of the data that research and public bodies rely on comes as tables, knowledge graphs and measurements over time. Before anyone can use it, it has to be found, and its description has to match what is in it. We work on both, and on learning from the data once it is found.

Part of this work is on the sea. In the RECOVER project we use marine monitoring data from the western Baltic to compare plankton communities and forecast how they change.

Themes
  • Dataset searchFinding tables and datasets by what is in them, and searching knowledge graphs.
  • MetadataWhether descriptions written by language models help people find data without getting it wrong.
  • Environmental dataComparing plankton communities and forecasting how they change, from monitoring data.
Papers
2026

PIPER: Content-Based Table Search via Profiling and LLM-Generated Pseudoqueries

Riccardo Terrenzi, Matteo Falconi, Serkan Ayvaz, Pierluigi Plebani

Database and Expert Systems Applications (DEXA 2026), Graz · LNCS, 197–212

2026

Why Neighborhoods Matter: Traversal Context and Provenance in Agentic GraphRAG

Riccardo Terrenzi, Maximilian von Zastrow, Serkan Ayvaz

GenAIK–NORA joint workshop at IJCAI 2026

2026

A Reference Architecture for Agentic Hybrid Retrieval in Dataset Search

Riccardo Terrenzi, Phongsakon Mark Konrad, Tim Lukas Adam, Serkan Ayvaz

Software Architecture and Machine Learning (SAML 2026) workshop at ICSA 2026, 542–548