O*NET Alternate Titles: Integrated Automation for Quality Control and Standardization

Published:

August 2025

Tags:
AI, Machine Learning, NLP   Occupational Taxonomy
Authors:

Brittany F. Crawford, Jiayi Liu, Daniel Barstow HumRRO

Phil Lewis National Center for O*NET Development

Summary:

Alternate Titles are a critical component of the O*NET database, providing users with job-level alternate or lay terminology linked to specific O*NET occupations. Alternate titles improve searchability and occupational understanding. These titles are collected from multiple sources, including write-in responses from job incumbents and occupational experts, data mining of employer job postings, and submissions from professional organizations and O*NET users (see: O*NET Alternate Titles Procedures).

The current manual review process for Alternate Titles write-in files can be time-consuming and requires substantial human effort, particularly where analysts need to exclude suggested titles that fail to meet established inclusion criteria. This paper describes the development of three automated systems to address these processing inefficiencies: (a) acronym and abbreviation standardization, (b) deletion of invalid or vague titles, and (c) exclusion of titles with occupational, contextual, or level mismatches.

Download: