Community Projects
Built from the NLx Research Hub data
Written with the NLx Research Hub data
Extracting O*NET Features from the NLx Corpus to Build Public Use Aggregate Labor Market Data
October 1, 2025
Authors: Stephen Meisenbacher, Svetlozar Nestorov, Peter Norlander
Data from online job postings are difficult to access and are not built in a standard or transparent manner. Data included in the standard taxonomy and occupational information database (O*NET) are updated infrequently and based on small survey samples. We adopt O*NET as a framework for building natural language processing tools that extract structured information from job postings. We publish the Job Ad Analysis Toolkit (JAAT), a collection of open-source tools built for this purpose, and demonstrate its reliability and accuracy in out-of-sample and LLM-as-judge testing. We extract more than 10 billion data points from more than 155 million online job ads provided by the National Labor Exchange (NLx) Research Hub, including O*NET tasks, occupation codes, tools, and technologies, as well as wages, skills, industry, and more features. We describe the construction of a dataset of occupation, state, and industry level features aggregated by monthly active jobs from 2015 – 2025. We illustrate the potential for research and future uses in education and workforce development.
Assessing Needs for Civilian STEM Talent in the Department of the Air Force
March 13, 2025
Authors: Sean Robson, Lisa M. Harrington, Kelly Atkinson, Alvin Moon, Nathan Thompson, Jonah Kushner, Grace Falgoust, Alice Nguyen, Barbara Bicksler, Joanna Zakzewski
This project examines the DAF’s needs for STEM talent in the civilian workforce, identifies potential gaps in technical competencies, describes options for closing technical skill gaps, and proposes strategies to better track the supply of and demand for STEM talent.
Because of the size and complexity of the DAF civilian workforce, our analysis centered on three case studies selected in consultation with the project sponsor: the Air Force Life Cycle Management Center, Digital Directorate (AFLCMC/HB; referred to as PEO Digital); the Deputy Chief of Staff, Air Force Futures; and Headquarters, Pacific Air Forces (HQ PACAF). The project team reviewed manpower and personnel data, conducted interviews with subject-matter experts (SMEs) in the case study organizations, analyzed relevant job announcement information, developed and tested two surveys, and conducted a literature scan of training and development opportunities.
Oregon 2024 Talent Assessment
May 10, 2024
Authors: Christiana McFarland, Annie Wyhof, Dylan Solden, Jacob Gottlieb, Gabrielle Christopher
To achieve equitable prosperity for all Oregonians and equip Oregon’s businesses and communities with the talent they need for success, the State of Oregon’s Higher Education Coordinating Commission (HECC) and Workforce and Talent Development Board (WTDB) collaborated with SRI International (SRI) to develop the Oregon 2024 Talent Assessment.
This assessment prioritizes four industry clusters that offer upward mobility and strong career pathways to workers in Oregon (Health Sciences, Information Technology & Analytical Instruments, Construction, and Wood Products Manufacturing), and two occupations that are essential to the functioning of Oregon’s economy (Childcare and Heavy and Tractor-Trailer Truck Drivers). To gain a comprehensive understanding of the opportunities and challenges within these industries and occupations, SRI examined labor market data and workforce and education system coordination, conducted an analysis of job postings, interviewed key partners in Oregon, and implemented a statewide employer survey.
Case Study: Assessing NLx Coverage of QCEW School Employers in Minnesota
and
Case Study: Assessing NLx Coverage of QCEW Construction Employers in Minnesota
March, 2024
Author: Amanda Rohrer
The NLx Research Hub data contains a massive amount of data from employers. They essentially self-select by using a job posting service and can move in and out of the universe. To meaningfully compare it to statistical sources of data, it’s necessary to have some sense of scope relative to a known universe of employers. Unfortunately, employers advertise jobs under other names and other hierarchies of ownership than they use to report their data to state and federal agencies. Job postings are also advertisements, not jobs. They sometimes run unfilled for several months and can be distributed widely (resulting in geographic duplication). They’re compiled from multiple sources – employers paying for a service and state job banks – and as a result coverage may vary between states. Making the comparison to established data programs requires custom matching of entities between the two sources. To see what kind of coverage the NLX data contains and the amount and type of work it would take to create a crosswalk between companies advertising jobs on NLX and the QCEW universe of liable employers, we used Schools (NAICS 6111), both public and private, as a test case for matching the two data sources.
Occupational Models from 42 Million Unstructured Job Postings
July 14, 2023
Authors: Nile Dixon, Marcelle Goggins, Ethan Ho, Mark Howison, Joe Long, Emma Northcott, Karen Shen, Carrie Yeats
Structuring jobs into occupations is the first step for analysis tasks in many fields of research, including economics and public health, as well as for practical applications like matching job seekers to available jobs. We present a data resource, derived with natural language processing techniques from over 42 million unstructured job postings in the National Labor Exchange, that empirically models the associations between occupation codes (estimated initially by the Standardized Occupation Coding for Computer-assisted Epidemiological Research method), skill keywords, job titles, and full-text job descriptions in the United States during the years 2019 and 2021. We model the probability that a job title is associated with an occupation code and that a job description is associated with skill keywords and occupation codes. Our models are openly available in the sockit python package, which can assign occupation codes to job titles, parse skills from and assign occupation codes to job postings and resumes, and estimate occupational similarity among job postings, resumes, and occupation codes.
Have a product that uses the NLx Research Hub data?
Fill out this form to help the community discover news ways to use the data!
Submissions are subject to NLx Research Hub approval before they are published on the NLx Research Hub website. It is the responsibility of the individual submitting to ensure they have the proper permissions to share the work.