Trustworthy AI, Safety & Governance
Privacy-preserving, explainable and accountable machine learning, and NLP for political and societal analysis.
Why it matters: Because AI used in health, migration and public discourse must be safe, private and open to scrutiny.
Featured events
- European patent — On-Air Split Federated Learning with Positive Labels — Intellectual property, 2024
- Founder & Chair — PoliticalNLP workshop series — Workshop leadership, 2022
- ITFLOWS — Horizon 2020 project (Beneficiary) — European project, 2020
Related projects
Publications (16)
- Towards Explainable Job Title Matching: Leveraging Semantic Textual Relatedness and Knowledge Graphs — RecSys in HR'25: 5th Workshop on Recommender Systems for Human Resources (ACM RecSys 2025), 2025
- Advancing Earth Observation: A Survey on AI-Powered Image Processing in Satellites — European Journal of Remote Sensing, 58(1), 2567921, 2025
- Proceedings of the Second Workshop on Natural Language Processing for Political Sciences @ LREC-COLING 2024 — LREC-COLING 2024, 2024
- Predicting Country Instability Using Bayesian Deep Learning and Random Forest — arXiv preprint, 2024
- Enabling All In-Edge Deep Learning: A Literature Review — IEEE Access, 11, 3431–3460, 2023
- Federated Split Learning with Only Positive Labels for Resource-Constrained IoT Environments — arXiv preprint arXiv:2307.13266, 2023
- Performance and Information Leakage in Splitfed Learning and Multi-Head Split Learning in Healthcare Data and Beyond — Methods and Protocols, 5(4), 60, 2022
- TransCasm: A Bilingual Corpus of Sarcastic Tweets — Proceedings of the LREC 2022 Workshop on NLP for Political Sciences, 2022
- Enabling Deep Learning for All-in Edge Paradigm — CoRR / arXiv preprint arXiv:2204.03326, 2022
- Proceedings of the LREC 2022 Workshop on Natural Language Processing for Political Sciences — LREC 2022 Workshop on NLP for Political Sciences, 2022
- D5.3 Manual for Software Prototype of Social Media Analysis Tool to Monitor Public Attitudes towards Migration — ITFLOWS Project Deliverable (H2020), 2022
- Splitfed Learning Without Client-Side Synchronization: Analyzing Client-Side Split Network Portion Size to Overall Performance — arXiv preprint arXiv:2109.09246, 2021
All 16 publications in this theme →
Milestones
- 2020 — ITFLOWS (H2020) — Beneficiary
- 2022 — Founded and chaired the PoliticalNLP workshop series
- 2024 — European patent — On-Air Split Federated Learning (WO/2024/227944)
See the full research agenda · collections · research journey.