Beyond the Data Gap: On the Relevance of a Multidisciplinary and Integrated Data-Driven Approach to Understanding Feminicide

Femicide is a pervasive and structurally rooted form of violence, yet it remains significantly underexplored using quantitative and computational approaches. Official data are often incomplete, aggregated, or inconsistent, limiting their analytical potential. This article applies an integrated data-driven approach that combines feminist theory, sociological insights, and machine learning techniques. Using microdata collected by the Italian feminist network Non Una di Meno, the study analyzes 434 cases of femicide recorded between 2020 and 2024. Following data cleaning and standardization, hierarchical cluster analysis identified three patterns based on victim-perpetrator relationships, motives, methods of killing, and demographic characteristics. The resulting clusters reveal distinct socio-relational dynamics of femicide. The findings highlight the value of feminist data activism and support greater integration of activist and institutional data to inform evidence-based prevention and policy.