Violence against children affects more than 1 billion around the globe, but it is a hidden phenomenon. Only 2 of every 100 cases of physical and sexual abuse are reported. Our project consists on using machine learning techniques to identify children exposed to violence and in this way enabling local governments to reach unprotected children. Our initial models suggest that around 70% of at-risk children could be identified, representing a 35 times improvement over the current level of detection.
SectorCivic Engagement & Policy