Hybrid Machine Learning Framework for Predicting Gender-Based Crimes: Integrating Socio-Economic and Regional Factors

This research addresses critical gaps in current hybrid machine learning models for risk assessment in gender-based crimes and crime prediction frameworks. Existing models focus on predicting crime occurrences but fail to specify the types and severity of future offenses. Additionally, these models lack comprehensive prevention strategies, overlook broader socio-economic factors influencing crime rates, and are often limited in their regional applicability. Using Spain's VioGen system as a case study, this paper proposes a new model framework that not only generalizes across different cultural and legal contexts but also integrates real-time socio-economic indicators and advanced data visualization tools for user-friendly, dynamic interaction. Furthermore, we consider the impact of post-pandemic conditions, such as increased isolation and stress, on gender-based violence patterns. By addressing these gaps, the research aims to enhance crime prediction accuracy, prevention strategies, and overall public awareness tools.