Offense Text Auto-Coder (OTAC) - Automated offense coder from offense charge text strings used for the BJS National Pretrial Reporting Program

Overview

Development stage

DeployedSource: c) Deployed – The use case is being actively authorized or utilized to support the functions or mission of an agency.

High-impact designation

Not high-impactSource: c) Not high-impact

Impact justification

Does not produce an output that serves as a principal basis for decisions or actions with legal, material, binding, or significant effect on any of the individuals or entities identified in OMB-25-21.

Start date

2024-01 (month precision)Source: 01/2024

Withheld from public reporting

NoSource: a) No

Mission

Topic area

Law Enforcement

Operational functions

No function tag matched Derived by keyword rules; see methodology

Problem the AI is intended to solve

The purpose of this tool is to improve description and comparability of offense charges across jurisdictions. When using justice administrative data from various jurisdictions (localities, states, and federal), the way offenses are described (i.e., the exact text strings used) vary greatly. For example, assault and battery may be spelled out, or abbreviated in novel ways such as A&B, A & B, A+B, A?B, battery & aslt, etc. This tool is used to facilitate grouping identical concepts under one common set of offense codes. The data that BJS makes available will be aggregated or deidentified.

Expected benefits

Improved comparisons of criminal justice data

System outputs

The output of this autocoder is common definitions for offense charge classification. It has been trained on an offense crosswalk for the BJS National Corrections Reporting Program (NCRP) (https://www.icpsr.umich.edu/files/NACJD/ncrp/Offense_Code_Crosswalk.xlsx) to convert plain-text offense descriptions into classifications routinely used by the NCRP.

Technology

AI classification

Natural language processingSource: Natural Language Processing: AI that processes, interprets, and shares information in human language.

System name(s)

Redacted for cybersecurity purposes.

Custom-developed code

Yes

Public source code

Not reported

Sourcing

How it was built

Developed in-houseSource: b) Developed in-house

Vendor (as reported)

Not reported

Vendors (standardized)

None on the standard list

A vendor is the supplier named by the agency. It does not identify the underlying model or AI technology.

Data and privacy

Involves PII

No

Privacy Impact Assessment

Not reported

Authorization to Operate

Yes

Demographic features

k) None of the above

Training and evaluation data

The case owner relied on DOJ AI governance practices to select and prepare data, as well as evaluate performance.

Federal Data Catalog

Not reported

Governance

0 of 8 minimum-practice questions answered. A blank answer means the agency reported nothing; it does not mean the practice is absent.

Potential impacts and how they were identified

Not reported