Domestic violence: a survey on norms and impacts, AI detection approaches, and existing datasets

Domestic violence: a survey on norms and impacts, AI detection approaches, and existing datasets

연구 분야: Infrastructure

학회: Knowledge and Information Systems

Domestic violence (DV) remains a pervasive issue globally, prompting extensive research to find effective solutions. This paper presents a comprehensive survey of significant DV research, focusing on societal norms, impacts, detection approaches, and existing datasets. We systematically reviewed and classified research articles from major repositories, including IEEE Xplore, ACM Digital Library, and Google Scholar. Key findings highlight the primary areas of focus in DV research, such as societal norms and perspectives, societal impact, and detection methods. The survey reveals that machine learning (ML) techniques are the most viable for detecting DV, with algorithms applied to various data forms, including text, audio, and video. We also provide brief explanations of related ML techniques and present recognized datasets, making the survey accessible to readers from diverse backgrounds. This paper may serve as a valuable reference for academicians, developers, and researchers, offering insights into the current state and future opportunities in DV research.