Enhancing Verification in iSHARE Data Spaces with Autonomous Self-Sovereign Digital IdentitiesVasilev, Nikolay Valentinov (Avtor) Lavbič, Dejan (Mentor) Data SpacesiSHARESelf-Sovereign IdentityMachine LearningHyperledger AriesVerifiable CredentialsVerifiable PresentationsSchema AlignmentLarge Language ModelsAs digital identity verification becomes more complex, the need for secure, decentralized solutions has never been greater. This is why we explored the potential of an autonomous Self-Sovereign Identity (SSI) within the iSHARE framework, utilizing advanced schema alignment with Large Language Models (LLMs) to enhance verification processes. By empowering entities to manage their credentials independently, this integration addresses key challenges in cross-organizational data exchange and identity verification. Our study results show that combining SSI and LLMs enables the generation of alternative Verifiable Presentation (VP) requests when initial requests fail, improving verification reliability in SSI systems. Additionally, SSI-enhanced iSHARE frameworks reduced processing time and enhanced privacy, security, scalability, flexibility, and error-handling capabilities in complex interactions. Our findings not only prove the effectiveness of SSI and LLM integration for decentralized identity verification in Data Spaces, but also reveal the potential of LLMs in advancing schema alignment tasks across diverse digital ecosystems. By fine-tuning LLM models with contextual embeddings, this research unlocks new opportunities for more adaptable and accurate mappings, offering future research fields in applying LLMs for schema alignment also in other data-sharing frameworks, identity management solutions, and dynamic environments where adaptive verification is critical.20252025-01-24 14:20:05Magistrsko delo/naloga166793VisID: 37029COBISS_ID: 224839683sl