ISWC 2026 · Half-Day Tutorial
Integrate CSV, JSON, XML, spreadsheets and more using SPARQL alone.
A growing community of researchers, practitioners, and tool builders has coalesced around Façade-X, a design philosophy and metamodel for querying heterogeneous data formats — CSV, JSON, XML, spreadsheets, and more — as knowledge graphs, using SPARQL as the sole integration language.
This tutorial brings that community to ISWC, disseminating the methods, technologies, and philosophy behind the approach. SPARQL Anything, the reference open-source implementation, serves as the primary teaching tool, while the tutorial also showcases independent implementations emerging in commercial platforms, reflecting the maturity of the approach. The W3C Data Façades Community Group, active since late 2025, provides the standardisation backdrop.
Integrating heterogeneous data formats remains one of the most persistent practical obstacles in knowledge graph construction. Conventional approaches rely on dedicated mapping languages and separate toolchains, adding complexity that often falls to domain experts rather than software engineers. Façade-X takes a different route: by exposing any format as a knowledge graph queryable with SPARQL alone, it removes the need for a second mapping language and keeps the entire workflow within familiar Semantic Web standards.
Understand the Façade-X design philosophy and its underlying metamodel.
Query heterogeneous data formats — CSV, JSON, XML and more — using SPARQL Anything.
Recognise real-world deployment patterns and data-integration strategies from production case studies.
Engage with the open theoretical questions that define the approach as an active research direction.
Appreciate the aims of the ongoing W3C standardisation effort and how to get involved.
Work through guided exercises on real datasets and bring your own questions to the SPARQL Anything clinic.
A half-day afternoon programme (14:10–18:00) built around the hands-on session, with all four themes represented.
| Time | Session | Presenter(s) |
|---|---|---|
| 14:10–14:50 | Slot 1 · The Problem and the Idea | |
| 14:10–14:30 | Introduction to Façade-X | Enrico & Luigi |
| 14:30–14:50 | Foundations: towards a theory of façade-based data access, with a short interactive modelling exercise | Enrico & Luigi |
| 14:50–16:10 | Slot 2 · Hands-on with SPARQL AnythingCore | |
| 14:50–15:20 | Live querying of CSV, JSON, and XML | Luigi |
| 15:20–15:45 | SPARQL Anything functions and magic properties | Enrico |
| 15:45–16:10 | Guided exercises with real datasets | Enrico & Luigi |
| 16:10–16:30 | Break · SPARQL Anything Clinic (bring your own data & questions) | All |
| 16:30–18:00 | Slot 3 · From Practice to Standards | |
| 16:30–16:50 | Façade-X deployment with TriplyDB platform (hands-on) | Martin |
| 16:50–17:10 | Railway infrastructure KG | Mathias |
| 17:10–17:30 | The Benefits of Decoupled Graph Construction. Lessons from the Public Procurement Data Space (slides) | Ivo |
| 17:30–17:40 | Pedagogical experience: teaching KG management with SPARQL Anything | Ryan |
| 17:40–18:00 | The W3C Façade-X Community Group: standardisation roadmap, community, and discussion | Enrico |
The tutorial deliberately avoids a lecture-only format: participants work through guided exercises on real datasets during the hands-on slot, bring their own data and questions to the SPARQL Anything clinic, and take part in an open discussion on community next steps.
Tutorial materials include slides for all sessions, Jupyter notebooks for the hands-on exercises, and SPARQL Anything documentation with worked examples.
All materials are openly licensed and will be made available here one week before the tutorial date.
The tutorial is application- and tool-focused, pitched at an intermediate level and targeted at knowledge graph engineers, data integration practitioners, Semantic Web researchers, and educators. Attendees are expected to have basic familiarity with SPARQL and at least one non-RDF data format (CSV, JSON, or XML). No prior knowledge of Façade-X or SPARQL Anything is assumed.
Associate Professor with 10+ years of AI research, a founding member of SPARQL Anything, and chair of the W3C Data Façades Community Group.
Senior Research Fellow specialising in KG construction for cultural heritage and smart cities, and a founding member of SPARQL Anything.
Software engineer leading SPARQL engine design and ELT pipeline development on the TriplyDB platform.
Independent railway and knowledge graph consultant with 17+ years in railway infrastructure data, ontology design, and SPARQL-based production deployments.
Associate Professor with 15 years teaching in information science; has taught KG management with SPARQL Anything over multiple semesters.
Independent consultant and trainer, PhD in Computer Science, 30+ years of experience and 20 years teaching, the last 8 on semantic technologies.