Privacy-Preserving Search, Information Retrieval, and AI over Solid Pods

Motivation

It can be argued that a more extensive use of data infrastructures based on Solid makes the requirements for
privacy-preserving search, information retrieval, and AI even stronger. At the same time, the better such
requirements are addressed and supported by Solid infrastructures, the wider the adoption of Solid can be.
AI over Solid Pods can not only support a plethora of decentralised applications but also leverage IoT and edge
computing resources. However, there are significant research challenges and trade-offs between performance and
privacy to be explored and addressed.

Goals

The workshop will foster discussion on the requirements and best practice for enabling information retrieval (IR)
and AI over decentralised data infrastructures based on Solid Servers and personal online datastores.
It will bring together researchers, practitioners, standardisation and policy experts to foster discussion and
consensus on the following:

1. The state of the art on algorithms and metadata for decentralised IR and AI.
2. Gaps in standardisation that need to be addressed.
3. Sectors that could benefit from IR and AI affordances in Solid, including sectors around Health and IoT.
4. Appropriate business models.

Topics of interest

• Decentralised IR/recommendation infrastructures and Personal Online Datastores (e.g., Solid, Dataspaces)
• Privacy-preserving IR, recommendation, and federated learning in decentralised settings
• Retrieval, ranking, and relevance modelling for distributed data sources
• Semantic interoperability, standards, and protocols for decentralised data integration and exchange
• Access control, consent management, and transparency mechanisms for user data
• Evaluation methodologies, metrics, and benchmarks for decentralised IR/recommendation systems
• User modelling, personalisation, and contextual relevance in decentralised environments

Call for Abstracts

We invite short abstracts aligned with the workshop topics listed above, covering ongoing or completed research
and/or position statements on privacy-preserving IR and AI over decentralised data infrastructures based on Solid.

Submissions must follow the Springer LNCS template and be submitted as a PDF.
Abstracts should be up to 2 pages (including references) and must be emailed to
george.popescu-craiova@kcl.ac.uk.

Submission deadline: 27 February 2026 (AoE)

Organization

• Mohammad Bahrani, University of Southampton
• George Popescu-Craiova, King’s College London

Chairs

• Thanassis Tiropanis, University of Southampton
• George Roussos, Birkbeck, University of London

Contact

Email:
George Popescu-Craiova (george.popescu-craiova@kcl.ac.uk)
and
Mohammad Bahrani (m.bahrani@soton.ac.uk)