Search results

Showing 1 – 4 of 4 results.
Curated

AI Enabled Community Supervision for Criminal Justice Services, 2020-2023 (ICPSR 38996)

Released/updated on: 2023-12-20

This project aimed to revolutionize the reentry process for justice-involved individuals (JII) by harnessing the power of artificial intelligence (AI) and advanced technologies. The centerpiece of the endeavor is the AI-based Support and Monitoring System, or AI-SMS, a cutting-edge platform designed to assist JII and their dedicated caseworkers in their journey to reintegrate seamlessly into the community. While the primary focus is on JII, the researchers recognize the critical role played by caseworkers-clinically trained individuals who facilitate the reentry process from a community perspective.

AI-SMS was conceived to be a multifaceted tool that provides case workers with early warning indicators of risky behavior and equips JII with the means and strategies to mitigate these risks, aligning with best practices in hybrid supervision. At its core, the system is committed to delivering personalized resources and opportunities to JII, complementing the support offered by caseworkers.

Curated
Simple Crosstabs

Applying Artificial Intelligence to Person-Based Policing Practices, 2019-2023 (ICPSR 39074)

Released/updated on: 2024-09-26
Time period: 2019-01-01--2023-01-01
In this project, the research team developed and evaluated an artificial intelligence (AI) tool using agent-based modeling methods for crime analysis and risk evaluation (CARE): CAREsim. The purpose of this tool was to improve the effectiveness of person-based patrol strategies, where police take preemptive actions upon selected high-risk individuals (determined based on factors known to police such as violent crime history) when predicted risks of committing crimes are high. CARESim was developed and tested with a simulated randomized controlled experiment within the jurisdiction of Hampton, Virginia. 240 high-risk individuals (120 in each group) were followed for a 12-month period, with the simulation lasting 23 months. The treatment group received additional crime analyses using the AI tool and more focused patrols, while the control group received analyses as usual and random patrols in the simulated environment. The tool was evaluated on a series of outcomes (e.g., number of crimes and arrests) comparing the control and treatment groups. This collection contains the simulated high-risk individual data (DS1) and the simulated crimes data (DS2) used for the experiment.
Curated
Simple Crosstabs

National Assessment of Demand Reduction Efforts, Part II: New Developments in the Primary Prevention of Sex Trafficking, [United States], 2021 (ICPSR 38928)

Released/updated on: 2026-01-14
Geographic coverage: United States

To combat prostitution and sex trafficking, criminal justice strategies and collaborative programs have emerged that focus on reducing consumer-level demand. From 2008 to 2012, the National Institute of Justice (NIJ) sponsored a study entitled "A National Overview of Prostitution and Sex Trafficking Demand Reduction Efforts" (referred to as Part I) that featured the systematic collection of information to determine the types and distribution of demand reduction tactics implemented throughout the United States. These efforts gave rise to a typology of law enforcement and community-based tactics identifying 12 different methods for deterring people (mostly men) from buying sex or which sanction those individuals who solicit sex acts. The essential product of that study was the Demand Forum website, launched in January 2013 by Abt Associates. In the years that followed, Demand Forum provided information about demand reduction interventions in the United States, and its content was updated and expanded through daily web searches and supplemented by periodic literature reviews or direct contact with a network of practitioners and other experts. During the website's first seven years of operation, it was viewed by more than 262,000 individuals from 179 countries and was used to shape policy and practice within the United States. However, innovations in the field, primarily new tactics using information technology (IT) to deter buyers and develop evidence to apprehend those actively seeking to purchase sex, have emerged since Demand Forum's launch.

The current study (referred to as Part II) builds upon the methodology and knowledge base of the initial study to keep the field informed of innovations and evolving responses to buyer behaviors and to continue to provide support for practice and policy. Beginning January 2021, the National Center on Sexual Exploitation (NCOSE), which now maintains Demand Forum, conducted a systematic assessment of current demand reduction tactics and created an expanded tactic typology to reflect recent innovations intended to reduce the demand that drives sex trafficking markets. The project also aimed to provide updated information and resources that could be used by practitioners. The methodology for identifying new information about existing tactics and their implementation in U.S. cities and counties featured a web-based survey distributed to more than 3,200 law enforcement agencies, more than 50 interviews with expert practitioners and survivors, searches of thousands of open source reports, reviews of the research and practice literature, and reviews of prostitution laws within all 50 states.

NOTE: Data collected from the survey and interviews during this project were intended to verify information about demand reduction tactics and were not meant for analysis. This collection is limited to the online survey data.

Curated
Restricted

Using Physician Behavioral Big Data for High Precision Fraud Prediction and Detection, United States, 2000-2019 (ICPSR 38811)

Released/updated on: 2025-12-02
Geographic coverage: United States
Time period: 2000-01-01--2020-12-31
This project used big data from non-clinical physician behavior. These include traffic violations, substance abuse, property ownership, stressors (e.g., bankruptcy and divorce), social media data, and other life events data. These variables, all based on public records, were used to construct a predictive model of Medicare fraud using machine learning techniques.