Policy-compliant data processing: RDF-based restrictions for data-protection
Download as PPTX, PDF0 likes182 views
The document outlines research on policy-compliant data processing, focusing on RDF-based restrictions for data protection and the ontology engineering process. It discusses various aspects of restriction modeling, the existing gaps in guidelines for the use of restrictions, and proposes a restriction modeling activity to support compliance assessments, particularly in privacy-aware contexts. The research aims to analyze ontology restrictions and improve support for knowledge engineers in creating relevant data structures.
Policy-compliant data processing: RDF-based restrictions for data-protection
- 1. Policy-compliant data processing: RDF-based restrictions for data-protection Sven Lieber, supervised by Anastasia Dimou and Ruben Verborgh ISWC 2019, Doctoral Consortium in Auckland
- 2. 2 Policy (App specific) Religion Email Person Sensitive 1 hasEmail maxCount Birthday Personal Violation if personal data used without informed consent Closed world constraints on data “Simple” modeling example: Person Consent 1 hasEmail maxCount 3
- 3. 3 Ontologies are carefully designed and a lot of design decisions need to be made
- 4. Ontologies are the result of a carefully executed process 4 Ontology or any knowledge representation in RDF Ontology Engineering process Existing ontologies Needs and Requirement s Concepts, Relationships, Restrictions
- 5. Currently unclear how needs and requirements are addressed by expressed restrictions 5 Ontology Ontology Engineering processNeeds and Requirement s Constraints
- 6. Unknown how restrictions are used in existing ontologies 6 OntologyOntology Engineering process Existing ontologies Needs and Requirement s Concepts, Relationships, Restrictions
- 7. Relevancy 7 Different use cases and level of formality Ontology with less and with more restrictions Annotation e.g. schema.org Reasoning e.g. pizza ontology
- 8. My research is relevant regarding the modeling of constraints when developing ontologies 8 Ontology with axioms RDFS, OWL Constraints on ontology concepts’ Instance data SHACL, ShEx
- 9. Restriction modeling in RDF as axioms, constraints or as part of policies is relevant to compliance checking 9 Privacy-aware data processing Compliance assessments to address this recognized need
- 10. Approach overview 10 Ontology Ontology Engineering process Existing ontologies Needs and Requirement s Constraints Concepts, Relationships, Restrictions 1) Inform 2) Compare existing activities and propose new activities 3) Apply on privacy use case
- 11. Related work: Restrictions can be modeled in a lot of different ways, but currently no guidelines when to use what 11 Existing ontologies Concepts, Relationships, Restrictions Closed world constraints (SHACL, ShEx) Open world axioms (OWL, RDFS) Restriction types in data quality
- 12. 12 Existing ontologies Concepts, Relationships, Restrictions How can we measure ontologies’ restriction use in a complete way? We can measure restriction use by obtaining statistics regarding restriction types and their different expressions. Research to understand the use of restrictions in RDF- based vocabularies
- 13. 13 Existing ontologies Concepts, Relationships, Restrictions Create statistics about the use of restriction types and expressions of ontologies from common ontology repositories Verify completeness of restriction type statistics by comparison with ground truth We want to create a statistical dataset about restriction use and evaluate its completeness
- 14. Preliminary Results indicate different use of Restrictions and implicit modeling patterns 14 Existing ontologies Concepts, Relationships, Restrictions Restriction use analysis of LOV and BioPortal MontoloStats - Ontology Modeling Statistics https://w3id.org/montolo Different restriction types
- 15. Second part of PhD covers the engineering process 15 OntologyOntology Engineering process Existing ontologies Needs and Requirement s Concepts, Relationships, Restrictions ✓
- 16. 16 Ontology Engineering processNeeds and Requirement s Related work: Ontology Engineering methodologies and Ontology Design patterns, but unclear how restriction modeling is supported Ontology Constraints
- 17. 17 How can we support users in restriction modeling involving axioms and constraints? We can define a restriction modeling activity which provides more support to model restrictions compared to existing activities. Extend existing ontology engineering activities to support restriction modeling Ontology Constraints
- 18. 18 How applicable is our restriction modeling activity in a privacy use case? Automatic privacy compliance checks by supporting the knowledge engineer in the creation of concepts, axioms, constraints and policies. Apply restriction modeling activity in privacy use case to support users in compliance assessments
- 19. Our approach to addressing the research 19 Define an Ontology Engineering activity to model restrictions based on existing best practices Apply activity in a different use cases Ontology Ontology Engineering processNeeds and Requirement s Constraints
- 20. User evaluations and case studies to evaluate our approach 20 Constraints User evaluations comparing restriction modeling activity with state-of-the-art activities Perform a case study in the privacy domain Ontology Ontology Engineering processNeeds and Requirement s
- 21. Preliminary Results cover proof-of-concepts regarding privacy-related annotations and a review of ontology engineering literature 21 Ontology Engineering processNeeds and Requirement s Ontology
- 22. Complement open world axioms with closed world constraints when systematically building knowledge Compliance assessments of privacy-aware data processing currently either axioms or constraints
- 23. Sven Lieber PhD researcher semantic intelligence E Sven.Lieber@ugent.be T +32 9 331 49 59 https://sven-lieber.org @SvenLieber Interested in - Ontology Engineering? - Shapes? - Privacy? Come and talk to me!