Domestic abuse affects people of every race, class, age, and nation. There is significant research on the prevalence and effects of domestic abuse; however, such research typically involves population-based surveys that have high financial costs. This work provides a qualitative analysis of domestic abuse using data collected from the social and news-aggregation website red-dit.com. We develop classifiers to detect submissions discussing domestic abuse, achieving accuracies of up to 92%, a substantial error reduction over its baseline. Analysis of the top features used in detecting abuse discourse provides insight into the dynamics of abusive relationships.
展开▼