Well-being is an important value for people's lives, and it is crucial for societal progress. Considering that well-being is a vague and multi-dimensional concept, it cannot be captured as a whole but through a set of health, socio-economic, safety, environmental, and political dimensions. The current Ph.D. thesis focuses on the safety dimension, and in particular on peace, which is an emerging challenge nowadays. Peace is the way out of inequity and violence, and its measurement is crucial, considering that the world is constantly under socio-economic, political, and military instability. Novel digital data streams and AI tools foster peace studies during the last years. Following this direction, we exploit information extracted from a new digital database called Global Data on Events, Location, and Tone (GDELT) to capture the Global Peace Index (GPI), a well-known official peace index. Applying predictive machine learning models, we demonstrate that news media attention from GDELT can be used as a proxy for measuring GPI at a higher frequency than the offcial yearly index cost- and time-efficiently. Additionally, we conduct variable importance analysis, and we use explainable AI techniques to understand better the models' behaviour, peace, and its determinants. This in-depth analysis highlights each country's profile and explains the predictions, prediction errors, and events that drive these errors. We believe that novel digital data exploited by researchers, policymakers, and non-governmental organisations, with data science tools as powerful as machine learning, could maximize the societal benefits and minimize the risks to peace and well-being as a whole.

Measuring well-being through novel digital data / Voukelatou, Vasiliki. - (2022 Jul 19).

Measuring well-being through novel digital data

VOUKELATOU, Vasiliki
2022-07-19

Abstract

Well-being is an important value for people's lives, and it is crucial for societal progress. Considering that well-being is a vague and multi-dimensional concept, it cannot be captured as a whole but through a set of health, socio-economic, safety, environmental, and political dimensions. The current Ph.D. thesis focuses on the safety dimension, and in particular on peace, which is an emerging challenge nowadays. Peace is the way out of inequity and violence, and its measurement is crucial, considering that the world is constantly under socio-economic, political, and military instability. Novel digital data streams and AI tools foster peace studies during the last years. Following this direction, we exploit information extracted from a new digital database called Global Data on Events, Location, and Tone (GDELT) to capture the Global Peace Index (GPI), a well-known official peace index. Applying predictive machine learning models, we demonstrate that news media attention from GDELT can be used as a proxy for measuring GPI at a higher frequency than the offcial yearly index cost- and time-efficiently. Additionally, we conduct variable importance analysis, and we use explainable AI techniques to understand better the models' behaviour, peace, and its determinants. This in-depth analysis highlights each country's profile and explains the predictions, prediction errors, and events that drive these errors. We believe that novel digital data exploited by researchers, policymakers, and non-governmental organisations, with data science tools as powerful as machine learning, could maximize the societal benefits and minimize the risks to peace and well-being as a whole.
Settore INF/01 - Informatica
Data Science
34
Articifial Intelligence (AI); AI for Social Good; Machine learning; Novel digital data; GDELT; news; Explainable AI; SHAP; Well-being; Peace; Global Peace Index
Scuola Normale Superiore
GIANNOTTI, Fosca
Pappalardo, Luca
File in questo prodotto:
File Dimensione Formato  
Thesis_Voukelatou_final.pdf

embargo fino al 18/07/2023

Tipologia: Tesi PhD
Licenza: Sola lettura
Dimensione 25.08 MB
Formato Adobe PDF
25.08 MB Adobe PDF   Visualizza/Apri   Richiedi una copia

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/11384/125822
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus ND
  • ???jsp.display-item.citation.isi??? ND
social impact