Harry Frank Guggenheim Foundation Research Grant Application

The proposed project is a continuation of a systematic analysis conducted by Yaksic (2015) where it was discovered that research on atypical homicide is complicated by variations in definition, sample size, data sources and collection procedures. Almost three decades ago, Kiger (1990) highlighted the limitations of employing then existing data to study the social problem of serial murder and called for the creation of new sources to allow for quantitative assessments that used empirical data. In response, atypical homicide researchers that previously operated in ‘information silos’ were encouraged to contribute data to the Radford Serial Killer Database Project to build a comprehensive record of serial homicide offending. An ongoing data sharing initiative was also organized with the Federal Bureau of Investigation’s Behavioral Analysis Unit 5 and the Homicide Investigation Tracking System of the Washington State Attorney General's Office. Statistical evidence generated from these databases enables analysts to disprove ingrained myths and stereotypes about serial murderers. The broad term ‘multiple-event murderer’ was adopted after a Delphi expert panel proposed reconceptualizing mass, spree and serial murderers together under atypical homicide. This approach permits researchers to pursue knowledge unburdened by disagreement on time intervals between homicides or discordance about the shared methods and motives within each subset of offenders.

The primary outcome of the proposed research will be the creation of an electronic surveillance tool to detect and track instances of crime indicative of burgeoning atypical homicide offenders. One aim of the study is to supplant the underutilized Violent Criminal Apprehension Program as only 5,000 entries were made to the system which equals less than half of expected submissions. To ensure optimal functionality of the dashboard interface, the root causes of serial violence and aggression must be understood and codified for the program to effectively pinpoint factors symptomatic of the modern day atypical killer. To this end, a team of raters from Northeastern University’s Atypical Homicide Research Group will objectively assign a probability score to each offense, ranking the likelihood that a crime is part of an offender’s overarching design. This objective will be completed in the early stages of training a computing cluster to recognize acts that are potentially part of a larger criminal pattern. The cluster will be tested and implemented at three law enforcement organization pilot sites which will feed offender information directly into future iterations of the dashboard. Locating instances of burgeoning atypical homicide offenders of any type will surely lead to the discovery of others because mass, spree and serial murderers advance through similar pathways. Mass homicide is an urgent problem in the modern world but efforts are dedicated to funding programs aimed at predicting instances of terrorism or addressing fatal shootings by police. Financing for the aforementioned methodologies is unavailable through other sources due to decreased attention and support. This project may be the commencement of how we will learn to control violence and aggression by intervening at strategic points over the larger scope of an offender’s activities and will improve upon the extant literature by introducing the field of atypical homicide research to concepts from the realm of computer science.