The XAI team participated in the Hybrid Human-AI Intelligence (HHAI) 2025 conference, held in Pisa from June 9th to 13th, 2025. The XAI project contributed to the organization of the event, facilitating engagement between the European research community and the local ecosystem on Human-Centered AI topics.
The conference provided an opportunity to present the project’s recent work, including the research line on Human-AI Coevolution, which examines the interaction between human decision-makers and AI systems.
Team members, including Simone Piaggesi, Clara Punzi, Samuele Tonati, and Giovanni Mauro, attended the sessions and presented new contributions regarding trustworthy and inclusive AI.
Co-organized Events
The XAI project played a direct role in the scientific program through the co-organization of specific workshops and tutorials:
AILS 2025: The First Workshop on AI, Labour and Society
Co-organized with the Carlo Azeglio Ciampi Institute of Advanced Studies and Scuola Normale Superiore, this workshop fostered an interdisciplinary dialogue between social scientists, economists, and AI experts. The sessions focused on the impact of AI, particularly Generative AI, on labor markets and society, addressing issues such as algorithmic management and the sustainable development of technology.
Tutorial: Assessing the impact of AI-driven recommenders on human-AI ecosystems
This tutorial offered a systematic overview of methodologies to measure the impact of recommender systems in socio-technical ecosystems. It covered domains including social media, online retail, and urban mapping, proposing a taxonomy for analyzing outcomes at individual and systemic levels, grounded in the framework of Human-AI Coevolution.
Contributions
The papers presented by the XAI team addressed challenges in public administration and natural language processing:
Trustworthy AI for Public Administration: The team presented a study on building a Retrieval-Augmented Generation (RAG) chatbot for the Italian public sector. This work evaluates embedding approaches to support information retrieval while complying with regulatory constraints (Mala et al., 2025).
Inclusivity in NLP: Another contribution proposed a multi-perspective framework for stance detection. By using soft labels, this approach models diversity and disagreement among human annotators to represent diverse viewpoints (Muscato et al., 2025).
The HHAI 2025 conference in Pisa served as a venue for the XAI project to discuss developments in hybrid intelligence and human-AI collaboration.
References
2025
Towards Building a Trustworthy RAG-Based Chatbot for the Italian Public Administration
Chandana Sree
Mala, Christian
Maio, Mattia
Proietti, Gizem
Gezici, Fosca
Giannotti, and
3 more authors
Building a Trustworthy Retrieval-Augmented Generation (RAG) chatbot for Italy’s public sector presents challenges that go beyond selecting an appropriate Large Language Model. A major issue is the retrieval phase, where Italian text embedders often underperform compared to English and multilingual counterparts, hindering precise identification and contextualization of critical information. Regulatory constraints further complicate matters by disallowing closed source or cloud based models, forcing reliance on on-premise or fully open source solutions that may not fully address the linguistic complexities of Italian documents. In our study, we evaluate three embedding approaches using a publicly available Italian dataset: a monolingual Italian approach, a translation based method leveraging English only embedders with backward reference mapping, and a multilingual framework applied to both original and translated texts. Our methodology involves chunking documents into coherent segments, embedding them in a high dimensional semantic space, and measuring retrieval accuracy via top-k similarity searches. Our results indicate that the translation based approach significantly improves retrieval performance over Italian specific models, suggesting that bilingual mapping can effectively address both domain specific challenges and regulatory constraints in developing RAG pipelines for public administration.
@inbook{MDP2025,author={Mala, Chandana Sree and di Maio, Christian and Proietti, Mattia and Gezici, Gizem and Giannotti, Fosca and Melacci, Stefano and Lenci, Alessandro and Gori, Marco},booktitle={HHAI 2025},doi={10.3233/faia250637},isbn={9781643686110},issn={1879-8314},line={3,5},month=sep,open_access={Gold},pages={196--204},publisher={IOS Press},title={Towards Building a Trustworthy RAG-Based Chatbot for the Italian Public Administration},visible_on_website={YES},year={2025}}
Embracing Diversity: A Multi-Perspective Approach with Soft Labels
Benedetta
Muscato, Praveen
Bushipaka, Gizem
Gezici, Lucia
Passaro, Fosca
Giannotti, and
1 more author
In subjective tasks like stance detection, diverse human perspectives are often simplified into a single ground truth through label aggregation i.e. majority voting, potentially marginalizing minority viewpoints. This paper presents a Multi-Perspective framework for stance detection that explicitly incorporates annotation diversity by using soft labels derived from both human and large language model (LLM) annotations. Building on a stance detection dataset focused on controversial topics, we augment it with document summaries and new LLM-generated labels. We then compare two approaches: a baseline using aggregated hard labels, and a multi-perspective model trained on disaggregated soft labels that capture annotation distributions. Our findings show that multi-perspective models consistently outperform traditional baselines (higher F1-scores), with lower model confidence, reflecting task subjectivity. This work highlights the importance of modeling disagreement and promotes a shift toward more inclusive, perspective-aware NLP systems.
@inbook{MBG2025,author={Muscato, Benedetta and Bushipaka, Praveen and Gezici, Gizem and Passaro, Lucia and Giannotti, Fosca and Cucinotta, Tommaso},booktitle={HHAI 2025},doi={10.3233/faia250654},isbn={9781643686110},issn={1879-8314},line={4,5},month=sep,open_access={Gold},pages={370--384},publisher={IOS Press},title={Embracing Diversity: A Multi-Perspective Approach with Soft Labels},visible_on_website={YES},year={2025}}
