Sheyla Leyva Sánchez

Sheyla Leyva SánchezDC5

Universidad Politécnica de Madrid · Open Models for AI/Data Risk Assessment

Bio

Sheyla Leyva Sánchez is a cuban computer scientist passionate about data and Artificial Intelligence. She completed her Master’s in Data Science at Universidad Politécnica de Madrid, where her curiosity led her to move beyond software development into scientific research.

Sheyla is a PhD researcher within the Marie Skłodowska-Curie Actions (MSCA) Doctoral Network, enrolled as a Doctoral Candidate at Universidad Politécnica de Madrid under the supervision of Prof. Víctor Rodríguez Doncel and Dr. María Poveda. Her PhD research examines the ethical and legal risks of AI systems within the framework of the EU AI Act and the Digital Services Act, with a focus on developing open semantic models and tools to support risk assessment and multi-stakeholder governance.

Research interests

AI Governance Ethical & Legal Risks Semantic Web Risk Assessment Models

Publications

  1. NORMA: A Semantic Framework for Legal Norm Representation from Annotated BPMN
    Leyva-Sánchez, S.; Poveda-Villalón, M.; Rodríguez-Doncel, V.; Quaranta, M.; Amantea, I.A.; Manab, M.A. — ISWC 2026 (to appear).
  2. Bias Inheritance in Model Genealogies: Toward Semantic Traceability for Compliance
    Leyva-Sánchez, S. — Doctoral Consortium, JURIX 2025, Turin, December 2025. PDF
  3. Extracting ODRL Policies from Business Process Models: A Graph Traversal Approach to Compliance-by-Extraction
    Manab, M.A.; Quaranta, M.; Amantea, I.A.; Leyva-Sánchez, S.; Rodríguez-Doncel, V. — 2nd International Workshop on ODRL and Beyond (OPAL 2026), collocated with ESWC 2026.
  4. DAOnt: A Formal Ontology for EU Data Act Compliance
    Leyva-Sánchez, S.; Linde, F.; Manab, M.A.; Poveda-Villalón, M.; Rodríguez-Doncel, V. — 3rd Conventicle on Artificial Intelligence Regulation and Safety, collocated with JURIX 2025, Turin, December 2025.

Deliverables

  1. D2.3 Tools for Data Governance
    WP2 · Lead: UPM — December 2026.
  2. D2.4 Risk Assessment and Management Models - v1
    WP2 · Lead: TCD — June 2026. DOI