Find Academic Supervisors

Rocco Tripodi is a Researcher at Ca' Foscari University of Venice's Department of Environmental Sciences, Computer Science and Statistics. He holds a Ph.D. in Computer Science (2015) from the same institution, focusing on game-theoretic models for Natural Language Processing (NLP). His roles include teaching courses such as Deep Learning for NLP, Introduction to Coding, and Database Systems. He has also served as an Assistant Professor at the University of Bologna (2021–2024) and collaborated with labs like Sapienza NLP and the European Centre for Living Technology (ECLT). His research spans NLP, machine learning, and computational linguistics, with a focus on lexical semantics, game theory applications, and ethical AI. Research Interests : His work addresses challenges in lexical semantics, including word sense disambiguation, semantic role labeling, and cross-lingual NLP. He explores game-theoretic frameworks for text analysis and develops tools for cultural heritage (e.g., historical text analysis). Recent projects include evaluating large language models (e.g., ChatGPT), detecting sensitive data in text, and tracing historical biases through diachronic embeddings. Publications : Tripodi's research trends focus on NLP fairness, multilingual models, and historical discourse analysis. Notable contributions include frameworks for abuse detection evaluation (AAA), multilingual semantic role labeling (UniteD-SRL), and diachronic analysis of antisemitic language in 19th-century French texts. His work bridges technical NLP advancements with societal impact, such as ethical AI and cultural heritage preservation. Labs & Collaborations : Active in interdisciplinary projects like ODYCCEUS and Polifonia, he collaborates with institutions to apply NLP to social sciences and humanities. His labs emphasize practical applications, such as transforming text into knowledge graphs (Text2AMR2FRED) and analyzing historical textiles using spectroscopic methods.

• Natural Language Processing
• Machine Learning
• Lexical Semantics
• +3 more

Paolo Torroni is an Associate Professor at the University of Bologna's Department of Computer Science - Science and Engineering since 2015. His research focuses on artificial intelligence, with emphasis on natural language processing, multi-agent systems, computational logics, and argumentation mining. He has authored over 180 scientific articles and directed the Language Technologies Lab, contributing to national and international projects. Since 2018, he holds full professor qualifications and has been a visiting fellow at the European University Institute. Education: PhD in Electronic and Computer Engineering (2002) from the University of Bologna. He conducted research at Imperial College London (2000-2003) and held research grants until 2005. Research Interests: His work spans AI applications in legal document analysis, privacy policies, healthcare diagnostics, and social good technologies. He develops tools like MARGOT for argument mining and CLAUDETTE for detecting unfair contract clauses. His recent studies include AI-driven chatbots for asylum seekers and explainable AI in Parkinson’s detection. Grants & Projects: Involved in numerous projects, including GDPR privacy policy analysis and EU-funded initiatives. He coordinates the international master’s in Artificial Intelligence and teaches courses like Natural Language Processing and Real-Time Operating Systems. Labs & Teams: Leads the Language Technologies Lab, collaborating with global research groups on multimodal argument mining, legal technologies, and ethical AI.

• Artificial Intelligence
• Natural Language Processing
• Multi-Agent Systems
• +5 more

Giulia Venturi is a Researcher at the Istituto di linguistica computazionale "Antonio Zampolli" (ILC), part of the Consiglio Nazionale delle Ricerche (CNR) in Italy. Her work focuses on computational linguistics, natural language processing, and corpus-based linguistic analysis, with particular expertise in Italian language processing. Dr. Venturi's research interests span computational linguistics, natural language processing, text simplification, linguistic profiling, corpus linguistics, parliamentary debates analysis, readability assessment, and large language models. Her work often combines theoretical linguistic insights with practical applications in NLP systems. She has made significant contributions to the development and analysis of linguistic resources, particularly for the Italian language. Her publication record demonstrates expertise in linguistic profiling of language models, corpus construction for parliamentary debates (notably the ParlaMint project), text simplification, and readability assessment. Her work often bridges theoretical linguistics with practical NLP applications, with a strong focus on the Italian language. Recent research has explored how linguistic knowledge can enhance encoder-decoder models, the evaluation of large language models through linguistic constraints, and the construction of high-quality treebanks for Italian parliamentary speeches. Dr. Venturi has been actively involved in major collaborative projects such as ParlaMint, which creates comparable multilingual corpora of parliamentary debates across Europe. Her work on readability assessment extends to healthcare communication, examining how institutional website readability affects access to services like abortion in Tuscany. She has also contributed to SemEval tasks focused on evaluating neural networks' understanding of presuppositional semantic knowledge. She collaborates extensively with researchers across Italy and Europe, particularly with Felice Dell'Orletta, Alessio Miaschi, and Dominique Brunato. Her work appears in major computational linguistics venues and contributes to important resources for the research community, including specialized corpora for text simplification and parliamentary discourse analysis.

• Computational Linguistics
• Natural Language Processing
• Text Simplification
• +5 more

Paola Perucchini is a Professor in the Department of Educational Sciences at the University of Rome III, specializing in child development, educational psychology, and environmental learning. Her work bridges early childhood education with interdisciplinary research on intercultural understanding, ADHD intervention, and sustainability pedagogy. Full Professor, Department of Educational Sciences Director of the Department of Educational Sciences Member of Academic Senate and Research Commission Her research focuses on reducing social biases, inclusive pedagogy, and the cognitive benefits of green environments. She pioneered the ODIS (Osservazione della DIScussione) framework for analyzing classroom communication and the Yesterday-Today-Tomorrow e-learning program for migration education. Recent publications (2023-2025) emphasize prosocial behavior promotion, intergroup coalition formation, and outdoor education's role in attention restoration. Earlier work (1997-2004) explores gestural development in autism and its implications for theory-of-mind cognition. She collaborates with the MuSEd – Mauro Laeng Museum of School and Education , integrating historical pedagogy with modern inclusive education strategies. Her teaching emphasizes socio-communicative teacher traits and remote learning adaptation, particularly in pandemic contexts.

• Educational Psychology
• Child Development
• Intercultural Education
• +3 more

Abelardo Carlos Martínez Lorenzo is a postdoctoral researcher at Sapienza University of Rome working under Professor Roberto Navigli in the Natural Language Processing group. His research focuses on cross-lingual semantics and information extraction, with specialization in semantic parsing and language models. His educational background includes: Software Engineering degree from Universidad de Málaga, Spain Erasmus Mundus Joint Master Degree in Big Data Management and Analytics from Université Libre de Bruxelles (Belgium), Universitat Politècnica de Catalunya (Spain), and Technische Universität Berlin (Germany) Martínez Lorenzo's research centers on overcoming language barriers through advanced semantic representation frameworks. His work develops efficient multilingual parsing systems that address data scarcity in low-resource languages, leveraging transformer architectures and novel linearization techniques. Key contributions include the BabelNet Meaning Representation (BMR) formalism and optimization methods for Abstract Meaning Representation (AMR) parsing, enabling robust cross-lingual semantic analysis without language-specific constraints. His publication record (2022-2024) reveals a consistent trajectory in enhancing AMR parser efficiency and multilingual capabilities. Notable innovations include CLAP's 80% reduction in computational time, cross-lingual alignment through transformer cross-attention, and ensemble methods that maintain structural integrity while improving SMATCH scores. These works collectively advance accessible, high-performance semantic analysis across linguistic boundaries. His scientific recognition includes: Marie Skłodowska-Curie Fellowship as Early Stage Researcher in the Knograph project (Horizon 2020) Funded through the prestigious Marie Curie grant, Martínez Lorenzo conducts research within the Knograph project framework while mentoring junior researchers in the Sapienza NLP group. His work demonstrates significant grant impact through open-source releases (CLAP, LeakDistill) that democratize access to semantic analysis tools. He actively contributes to the Sapienza Natural Language Processing group led by Roberto Navigli, which specializes in knowledge-based multilingual NLP systems. The group maintains strong collaborations through European research initiatives including Horizon 2020 projects, with emphasis on creating interlingual semantic resources and scalable parsing frameworks.

• Natural Language Processing
• Cross-lingual Semantics
• Information Extraction
• +3 more

Giovanni Trappolini is an Assistant Professor in the Department of Computer, Control, and Management Engineering at Sapienza University of Rome, where he conducts research in the RSTLess Lab under the supervision of Prof. Fabrizio Silvestri. He earned his PhD in Machine Learning from Sapienza in 2022, focusing on Geometric Deep Learning under Prof. Emanuele Rodolà. Trappolini has collaborated with leading institutions such as Stanford, Technion, Meta, Amazon, TII, and the University of Pisa, contributing to cutting-edge research in AI and information retrieval. His research interests include machine learning, deep learning, geometric deep learning, multimodal AI, information retrieval, and large language models. He has pioneered work on Retrieval-Augmented Generation (RAG), Neural Databases, and the development of Fauno, the leading Italian LLM. His teaching experience includes courses on Python for Data Science, Advanced Data Mining, and Statistical Learning at both Sapienza and Luiss Guido Carli. Trappolini's recent publications span top-tier venues like SIGIR, NeurIPS, ECCV, and IEEE Transactions, with a focus on graph neural networks, federated learning, anomaly detection in 5G networks, and multimodal systems. His work shows a strong trend toward robust, privacy-preserving, and multimodal AI systems with real-world applications in language, vision, and network security. Sapienza honor graduate IELTS 8.0 He has advised several teaching roles and delivered courses on Python, data science, and statistical learning. His research is supported through collaborations with industry leaders like Meta and Amazon, and he leads innovative projects such as Fauno and Neural Databases. Trappolini is actively involved in the AI research community and continues to push the boundaries of language and retrieval systems. He is a core member of the RSTLess Lab and has been instrumental in advancing geometric and multimodal deep learning at Sapienza University. His work bridges theoretical innovation with practical implementation, particularly in the domains of Italian language modeling and secure, efficient information retrieval systems.

• Machine Learning
• Deep Learning
• Geometric Deep Learning
• +9 more