"The Intimate Partner Violence Criminal Career:Predicting Escalation Patterns and Turning Points Across Seven Million People Using AI

本研究将亲密伴侣暴力（IPV）与犯罪职业研究相结合，探讨了两个犯罪职业概念在IPV现象中的应用：升级和转折点。具体而言，本研究通过确定(1)预测IPV发生“拐点”的因素，(2)IPV二元组合的升级模式，以及预测未来升级模式的呈现因素，对现有的学术成果做出了贡献。特别是，调查升级模式是有用的，因为这些模式之间的明显区别可能反映了IPV犯罪者和受害者之间的异质性。因此，检查导致这种异质性的因素可以支持设计用于解释和预防IPV的理论和应用模型的发展。最后，本研究对大数据的使用促进了刑事司法实践中一个快速出现的趋势，对如何监管IPV具有重要而新颖的意义。正如Boyd和Crawford（2012，第663页）所言，大数据的定义是“搜索、聚合和交叉引用大型数据集的能力”，以一种为最终用户提供价值的方式，实现了本研究的目标。

This study integrates Intimate Partner Violence (IPV) and criminal careers research by exploring the application of two criminal career concepts within the phenomenon of IPV: escalation and turning points. Specifically, this study contributes to existing scholarship by ascertaining (1) the factors predictive of the 'turning point' into IPV perpetration and (2) the escalation patterns of IPV dyads, and the presenting factors that are predictive of future escalation patterns. In particular, it is useful to investigate escalation patterns since demonstrated distinctions between these patterns may reflect heterogeneity among IPV offenders and victims.Examining the factors that contribute to such heterogeneity can thus support the development of theoretical and applied models designed to account for and prevent IPV. Finally, this study's use of big data contributes to a fast-emerging trend within criminal justice practice, with important and novel implications for how IPV is policed. As Boyd and Crawford (2012, pg. 663) argue, big data is defined by its 'capacity to search, aggregate, and cross-reference large data sets' in a way that provides value to the end-user, objectives met by this study.