News Story

From lab to standard: KMi research behind the new W3C Data Façades community group

Thursday 9 Apr 2026

KMi is proud to share a milestone for one of its research lines in knowledge graphs engineering: the launch of the Data Façades Community Group at the W3C, the international standards body for Web technologies. The group’s formation marks the first step towards standardisation. It is an exciting moment, and one that has been a long time in the making.

The idea

Any practitioner working with real-world data knows the frustration: data lives in CSV files, JSON APIs, XML feeds, spreadsheets, relational databases, and each format demands its own transformation pipeline. For knowledge graph engineers, this has long meant writing complex mappings or ad-hoc code, requiring familiarity with both source formats and dedicated mapping languages.

Façade-X was conceived as a simpler alternative: a minimalist meta-model that treats any data source, regardless of format, as a uniform structure of containers, slots, and values in RDF. Rather than learning a new transformation language, engineers can query any data format directly using the SPARQL they already know. The idea was first published in “Facade-X: an opinionated approach to SPARQL anything“ by Enrico Daga (KMi) and Luigi Asprino (ISTC-CNR), together with Paul Mulholland (KMi) and Aldo Gangemi (ISTC-CNR). A follow-up article, “Knowledge Graph Construction with a Façade“, extended the approach to the full knowledge graph construction workflow.

The reference implementation, SPARQL Anything, demonstrated practical feasibility across JSON, CSV, XML, HTML, spreadsheets, and more. At KMi, the work has also involved Jason Carvalho, Marco Ratta, and Paul Warren, who have contributed to research, development, and evaluation.

The Community Group

The Data Façades Community Group aims to make Façade-X a vendor-neutral, interoperable technology that any developer or organisation can implement, decoupling it from the original SPARQL Anything codebase. Work happens openly on GitHub, with meeting minutes publicly archived. Current specification drafts, covering the formal metamodel and RDF vocabulary, are available at w3c-facade-x.github.io/facade-x-specs.

Equally important has been the early engagement of industry experts who helped shape the approach from outside academia: Justin Dowdy, Ivo Velitchkov, and Mathias Vanden Auweele have all brought practitioner perspectives that have been invaluable in grounding the work in real use cases.

Industry adoption

Encouragingly, companies are already implementing Façade-X in their products. Maplib (Data Treehouse) is a high-performance Rust-based knowledge graph library for Python that integrates the Façade-X approach for heterogeneous data access.Triply, a leading knowledge graph infrastructure company, has also begun engaging with the approach for their TriplyDB platform. Their participation in the Community Group brings production-scale implementation experience that will be essential for developing specifications that work in the real world.

New theoretical work

Alongside the standardisation effort, Luigi Asprino and Enrico Daga have published “Towards a theory of Façade-X data access: satisfiability of SPARQL basic graph patterns”, forthcoming in the Special Issue: Data Management for (Knowledge) Graphs of Transactions on Graph Data and Knowledge. The paper studies which SPARQL queries are satisfiable when evaluated over Façade-X data sources, a question with direct implications for query optimisation and the development of more efficient data integration systems.

Looking ahead

The Data Façades Community Group is a beginning, not an endpoint, the plan is to develop mature specifications and progress to a full W3C Working Group. With solid research foundations, growing industry adoption, and an active open community, the pieces are in place.

If you work with heterogeneous data and knowledge graphs, we invite you to follow the group’s progress, or join the conversation.

Useful links: