Presentations

  1. Private Federated Learning Knowledge Graph Representation Learning at Lab Meeting, University of Insubria, Varese, Italy (10/2022)

    This presentation introduces my investigation on the privacy protection techniques for learning knowledge graph representation under federated learning settings by using differential privacy.

  2. Time-Aware Anonymization of Knowledge Graphs at the Cyber Security Competence for Research and Innovation Project (CONCORDIA) Meeting, Munich, Germany (06/2022)

    This presentation introduces my k-anonymity technique protecting data owners from identity and attribute leakage even though adversaries have access to all published anonymized versions of a knowledge graph.

  3. Transparency in Proximity Advertising Campaigns at the Cyber Security Competence for Research and Innovation Project (CONCORDIA) Meeting, Munich, Germany (06/2022)

    This presentation introduces my blockchain-based platform ensuring the transparency of marketing campaign effectiveness measurement and privacy of participants (marketers, publishers, and customers).

  4. Differential Privacy: Foundation, Applications, and Challenges at Lab Meeting, University of Insubria, Varese, Italy (09/2021)

    This presentation covers basic concepts of differential privacy.

  5. Privacy-Preserving Sequential Publishing of Knowledge Graphs at the 37th IEEE International Conference on Data Engineering, Chania, Greece (Virtual) (04/2021)

    This presentation introduces the sequential anonymization approach for protecting users from identity leakage when adversaries have access to w continuous anonymized versions of a knowledge graph.

  6. Adversarial Machine Learning at Lab Meeting, University of Insubria, Varese, Italy (11/2020)

    This presentation introduces some adversarial machine learning attacks.

  7. Cluster-based Anonymization of Knowledge Graphs at the 18th International Conference on Applied Cryptography and Network Security, Rome, Italy (Virtual) (10/2020)

    This presentation introduces the first anonymization approach for knowledge graphs.