Assessing the Impact of Covid-19 on Crime Rates in Italy Using Machine Learning

The Covid-19 pandemic has had wide-ranging socio-economic consequences whose long-term impact might prove difficult to assess. One such area is crime. While studies are available on the effect of Covid-19 on crime rates in Italy, we feel this is a highly impactful research area that has not received enough attention, and robust analyses based on ML may strongly contribute to our understanding of the phenomenon and have a positive, lasting impact.

To leverage ML and AI tools to further our understanding of Covid-19’s effects on crime. This would include both the “short-term” impacts and “long-term” effects that might to this day not be fully understood. As such, problems to be addressed may include: 

• Have, e.g., lockdowns reduced or increased crime rates overall? 
• In which areas and regions have these changes proved most dramatic? 
• Have certain specific types of crimes narrowed in frequency while others (such as domestic violence) have spread further both at the time of the pandemic and ot date? 
• How has the “geography” of crime changed overall? 
• What specific factors, if any, can we isolate that explain these changes? 
• Are there significant differences in how they have affected urban rather than rural areas? 
• What can we predict about these trends in the near- and medium-term?

1. To identify the impact of Covid-19 and Covid-19 containment policies such as lockdowns on crime in Italy
2. Build ML models to identify, measure and predict such factors
3. Predict which trends are likely to last and which are more transient

1. Contribute to local challenges
2. Educate and empower local talent while contributing to the real-world local problems solutions
3. Flag results to the local policymakers to help them on their data-driven decision-making