Predictive modeling and spatiotemporal analysis of gender-based violence in Rajasthan, India.

Crimes against women, including dowry deaths, miscarriage, and cruelty by husband or in-laws, remain a pervasive social issue in India. This study examines the spatiotemporal patterns and spatial autocorrelation of these crimes in Rajasthan, India, before, during, and after the COVID-19 lockdown. Leveraging data from the National Crime Records Bureau from 2018 to 2022, we employed spatial scan statistics to identify high-intensity hotspots and analyze their evolution. For spatial autocorrelation, Moran's I Index was evaluated and a Logistic regression model was employed to identify factors affecting hotspot formation. Ajmer emerged as a major hotspot during the lockdown, while Bhilwara, Chittorgarh, Rajsamand, Bundi, Kota, and Ajmer remained hotspots before and after the lockdown. Spatiotemporal analysis revealed 2020 lacked the highest likelihood ratio for hotspot formation, and Moran's I showed positive spatial autocorrelation pre- and post-COVID-19, with lower values during the lockdown. The modeling identified the factors associated with hotspot formation as the poverty rate, per capita income, unemployment, substance use by men, early pregnancy, and disability in females. These findings highlight the complex nature of the event, shaped by socioeconomic and demographic factors, and offer insights to guide targeted policies and resource allocation to combat the menace of crime against women in Rajasthan.