The Gun Violence Database

TL;DR

The paper introduces the Gun Violence Database (GVDB), a comprehensive, crowdsourced, and NLP-automated resource that aims to facilitate objective research and discussion on gun violence in the U.S.

Contribution

It presents a large-scale, publicly accessible database built from news reports, with efforts to automate data collection using advanced NLP techniques.

Findings

Successfully compiled a large database of gun violence incidents.

Implemented NLP methods for automated data extraction.

Facilitated objective, data-driven research on gun violence.

Abstract

We describe the Gun Violence Database (GVDB), a large and growing database of gun violence incidents in the United States. The GVDB is built from the detailed information found in local news reports about gun violence, and is constructed via a large-scale crowdsourced annotation effort through our web site, http://gun-violence.org/. We argue that centralized and publicly available data about gun violence can facilitate scientific, fact-based discussion about a topic that is often dominated by politics and emotion. We describe our efforts to automate the construction of the database using state-of-the-art natural language processing (NLP) technologies, eventually enabling a fully-automated, highly-scalable resource for research on this important public health problem.