Maria Paola Priola
About Me
I hold a PhD in Computational Statistics from the University of Cagliari (Italy), with a dissertation titled “Essays on Data Frameworks and Sustainable AI for Public Health”. My research lies at the intersection of data engineering, artificial intelligence, and statistical evaluation, with a focus on developing reliable, interpretable, and sustainable data-driven solutions for healthcare.
My work explores three main research directions: (i) improving the trustworthiness of Large Language Models through Retrieval-Augmented Generation (RAG) and novel evaluation frameworks such as the Negative Missing Information Scoring System (NMISS); (ii) designing data-centric infrastructures, including a multimodal healthcare lakehouse (MEDITA) for integrating and analyzing heterogeneous public health data; and (iii) advancing Green AI practices by assessing trade-offs between predictive performance, computational efficiency, and environmental impact in machine learning models.
I also hold a Master’s degree in Data Science and both a Master’s and a Bachelor’s degree in Economics. Before returning to academia, I worked for over seven years in the private sector. At Prometeia (Bologna, Italy), I served as an econometrician and data scientist, contributing to projects involving natural language processing, machine learning, and text/audio analysis for central banks and monetary policy. I coordinated junior staff and delivered training on Excel and VBA.
My academic and professional path includes several international experiences: first MSc thesis in Buenos Aires (IAE Business School), an Erasmus exchange at WU Wien (Vienna University of Economics and Business), a finance traineeship in Bucharest at Confindustria Romania through the Erasmus Placement program, and additional periods abroad in Scotland and Ireland.
Fluent in English and Spanish, Bloomberg and Reuters user. Proficient in Python, R, SQL, Excel/VBA, Windows and Linux environments.
I am also the author of Tutto ciò che ci accomuna, a book published as part of a charity project supporting mental health awareness. The initiative includes Voices and Words for Mental Health, a musical tribute to international artists who have experienced mental health challenges.
In my spare time, I enjoy reading, writing, and playing music — I am also a guitarist. I’m passionate about exploring different cultures, and I love sports such as running, snowboarding, and trekking.
Publications
-
Balancing Performance and Environmental Efficiency: a Multiclass Classification Study of Textual Data
Electronic Journal of Applied Statistical Analysis, 2025. Volume 18, Issue 2: 238–263.
Authors: Maria Paola Priola, Maurizio Romano
DOI: 10.1285/i20705948v18n2p238 -
Unmasking Climate Risk in Earnings Calls: Traversing Storms and Fire with a Taxonomy-GPT Prompting Approach
Proceedings of the Conference on Green Finance Research Advances, CEUR Workshop Proceedings, Vol. 3762, 2024. Paper 54.
Authors: Michele Cimino, Annalisa Molino, Maria Paola Priola, Lorenzo Prosperi, Lea Zicchino
Available at: CEUR-WS Vol. 3762, Paper 54 -
Multi-class text classification of news data
Proceedings of the Statistics and Data Science 2024 Conference. New perspectives on Statistics and Data Science.
Authors: Maurizio Romano, Maria Paola Priola
View on handle.net -
The informative value of central banks talks: a topic model application to sentiment analysis
Data Science in Finance and Economics, 2022, Volume 2, Issue 3: 181–204.
Authors: Maria Paola Priola, Annalisa Molino, Giacomo Tizzanini, Lea Zicchino
Presented at: Advanced analytics: new methods and applications for macroeconomic policy, Bank of England, London – July 2022
DOI: 10.3934/dsfe.2022009 -
Fed Rate Hikes and the Effects on Turkish Banking System
Turkish Economic Association International Conference on Economics ICE-TEA, 2021.
Authors: Salvatore Perdichizzi, Maria Paola Priola, Lorenzo Prosperi, Giacomo Tizzanini, Lea Zicchino
DOI: 10.2139/ssrn.3309671
Working Papers
-
Unmasking Physical Risk with STORM: An NLP Framework for Natural Hazard Assessment in Earnings Calls
SSRN, 2025.
Authors: Michele Cimino, Annalisa Molino, Maria Paola Priola, Lorenzo Prosperi, Lea Zicchino
DOI: https://dx.doi.org/10.2139/ssrn.4879378 -
Addressing Hallucinations with RAG and NMISS in Italian Healthcare LLM Chatbots
Preprint, 2024.
Authors: Maria Paola Priola
View on arXiv -
Measuring Central Banks’ Sentiment and its Spillover Effects with a Network Approach
SSRN, 2021.
Authors: Maria Paola Priola, Piero Lorenzini, Giacomo Tizzanini, Lea Zicchino
Presented at: Banca d'Italia and Federal Reserve Board Joint Conference on Nontraditional Data & Statistical Learning - November, 2020
DOI: 10.2139/ssrn.3764004
Conferences
-
ITADATA 2025: The 4th Italian Conference on Big Data and Data Science
Turin, Italy – September 9–11, 2025
Presented the paper: From Data Sources to User Screens: a Multimodal hEalth Data Lakehouse for ITAly (MEDITA) -
International Conference on Data Science, Technology and Applications (DATA)
(scientific peer-reviewed conference)
Presented the work: A Multimodal hEalth Data warehouse for ITAly (MEDITA) -
Advanced Analytics: New Methods and Applications for Macroeconomic Policy
Bank of England, London, United Kingdom – July 2022
Presented the paper: The informative value of central banks’ talks: a topic model application to sentiment analysis -
Banca d’Italia and Federal Reserve Board Joint Conference on “Nontraditional Data & Statistical Learning with Applications to Macroeconomics”
2020
Presented the paper: Measuring Central Banks’ Sentiment and its Spillover Effects with a Network Approach