Sources

1. U.S. Department of Labor, Employment and Training Administration, Office of Foreign Labor Certification. LCA Programs (H-1B, H-1B1, E-3) Disclosure Data, FY2025 Q1 to Q4. 2025. dol.gov/agencies/eta/foreign-labor/performance. dol.gov refuses scripted downloads, so the four files were read from the Internet Archive's copies of the same URLs (captured February, August, August and December 2025).
2. Each company's own public job feed, read weekly; the feed for each is listed in the README. Postings are the companies' words, linked from the map.
3. National Center for O*NET Development. O*NET 31.0 Database: Technology Skills. U.S. Department of Labor, 2026. onetcenter.org/database.html
4. European Commission. ESCO (European Skills, Competences, Qualifications and Occupations) v1.2.1. 2025. esco.ec.europa.eu
5. U.S. Census Bureau. City and Town Population Totals: 2020-2024 (SUB-EST2024) and 2024 Gazetteer Files: Places. census.gov
6. Mike Bostock. us-atlas 3.0.1 (from the Census Bureau's cartographic boundary files), with d3 7.9.0 and topojson-client 3.1.0. github.com/topojson/us-atlas
7. Qwen Team. Qwen3 Embedding (Qwen3-Embedding-8B). Alibaba, 2025. huggingface.co/Qwen/Qwen3-Embedding-8B. Topic names by Qwen3.6 run locally in Ollama.
8. Leland McInnes, John Healy and James Melville. UMAP: Uniform Manifold Approximation and Projection for Dimension Reduction. arXiv:1802.03426, 2018.
9. Burt L. Monroe, Michael P. Colaresi and Kevin M. Quinn. "Fightin' Words: Lexical Feature Selection and Evaluation for Identifying the Content of Political Conflict." Political Analysis 16(4), 2008. The log-odds method behind "what sets each apart".
10. U.S. Bureau of Labor Statistics. Occupational Employment and Wage Statistics, May 2025, national estimates (oesm25nat.zip), 15-2051 Data Scientists. 2026. bls.gov/oes/tables.htm

Code and data: apps/web/data-science-jobs/. The snapshot script is modelled on What the AI labs are hiring for.