Validation of a Small Language Model for DSM-5 Substance Category Classification in Child Welfare Records

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Perron, Brian E., Stoll, Dragan, Victor, Bryan G., Qia, Zia, Jud, Andreas, Ryan, Joseph P.
Format: Preprint
Published: 2026
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866918377912532992
author Perron, Brian E.
Stoll, Dragan
Victor, Bryan G.
Qia, Zia
Jud, Andreas
Ryan, Joseph P.
author_facet Perron, Brian E.
Stoll, Dragan
Victor, Bryan G.
Qia, Zia
Jud, Andreas
Ryan, Joseph P.
contents Background: Recent studies have demonstrated that large language models (LLMs) can perform binary classification tasks on child welfare narratives, detecting the presence or absence of constructs such as substance-related problems, domestic violence, and firearms involvement. Whether smaller, locally deployable models can move beyond binary detection to classify specific substance types from these narratives remains untested. Objective: To validate a locally hosted LLM classifier for identifying specific substance types aligned with DSM-5 categories in child welfare investigation narratives. Methods: A locally hosted 20-billion-parameter LLM classified child maltreatment investigation narratives from a Midwestern U.S. state. Records previously identified as containing substance-related problems were passed to a second classification stage targeting seven DSM-5 substance categories. Expert human review of 900 stratified cases assessed classification precision, recall, and inter-method reliability (Cohen's kappa). Test-retest stability was evaluated using approximately 15,000 independently classified records. Results: Five substance categories achieved almost perfect inter-method agreement (kappa = 0.94-1.00): alcohol, cannabis, opioid, stimulant, and sedative/hypnotic/anxiolytic. Classification precision ranged from 92% to 100% for these categories. Two low-prevalence categories (hallucinogen, inhalant) performed poorly. Test-retest agreement ranged from 92.1% to 99.1% across the seven categories. Conclusions: A small, locally hosted LLM can reliably classify substance types from child welfare administrative text, extending prior work on binary classification to multi-label substance identification.
format Preprint
id arxiv_https___arxiv_org_abs_2603_06836
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Validation of a Small Language Model for DSM-5 Substance Category Classification in Child Welfare Records
Perron, Brian E.
Stoll, Dragan
Victor, Bryan G.
Qia, Zia
Jud, Andreas
Ryan, Joseph P.
Computation and Language
General Literature
Background: Recent studies have demonstrated that large language models (LLMs) can perform binary classification tasks on child welfare narratives, detecting the presence or absence of constructs such as substance-related problems, domestic violence, and firearms involvement. Whether smaller, locally deployable models can move beyond binary detection to classify specific substance types from these narratives remains untested. Objective: To validate a locally hosted LLM classifier for identifying specific substance types aligned with DSM-5 categories in child welfare investigation narratives. Methods: A locally hosted 20-billion-parameter LLM classified child maltreatment investigation narratives from a Midwestern U.S. state. Records previously identified as containing substance-related problems were passed to a second classification stage targeting seven DSM-5 substance categories. Expert human review of 900 stratified cases assessed classification precision, recall, and inter-method reliability (Cohen's kappa). Test-retest stability was evaluated using approximately 15,000 independently classified records. Results: Five substance categories achieved almost perfect inter-method agreement (kappa = 0.94-1.00): alcohol, cannabis, opioid, stimulant, and sedative/hypnotic/anxiolytic. Classification precision ranged from 92% to 100% for these categories. Two low-prevalence categories (hallucinogen, inhalant) performed poorly. Test-retest agreement ranged from 92.1% to 99.1% across the seven categories. Conclusions: A small, locally hosted LLM can reliably classify substance types from child welfare administrative text, extending prior work on binary classification to multi-label substance identification.
title Validation of a Small Language Model for DSM-5 Substance Category Classification in Child Welfare Records
topic Computation and Language
General Literature
url https://arxiv.org/abs/2603.06836