{"id":"W4410884811","doi":"10.1257/pandp.20251104","title":"From Online Job Postings to Economic Insights: A Machine Learning Approach to Structuring Naturally Occurring Data","year":2025,"lang":"en","type":"article","venue":"AEA Papers and Proceedings","topic":"Labor market dynamics and wage inequality","field":"Economics, Econometrics and Finance","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Bank of Canada","funders":"","keywords":"Structuring; Data science; Computer science; Machine learning; Artificial intelligence; Business","routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":true,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":["metaepi_narrow"],"consensus_categories":[],"category_scores_codex":[0.0004347613,0.0002609793,0.0004687834,0.0002698403,0.0002344386,0.0003506368,0.0006943793,0.0001187081,0.00003233424],"category_scores_gemma":[0.0003096149,0.0002732982,0.00004799843,0.0002932256,0.00002687687,0.0003783211,0.0009181551,0.0003468098,0.00001525712],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001162621,"about_ca_system_score_gemma":0.00002302388,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002632079,"about_ca_topic_score_gemma":0.0001699143,"domain_scores_codex":[0.9980285,0.00000611274,0.0005770048,0.001003134,0.00004303321,0.0003422852],"domain_scores_gemma":[0.9992502,0.00006260528,0.0001844435,0.0002868578,0.00003842538,0.0001774954],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"observational","study_design_scores_codex":[0.0002792078,0.0001211667,0.6525257,0.000382933,0.000433672,0.000005213941,0.009161445,0.0005206917,0.001752548,0.3137047,0.0009382232,0.02017461],"study_design_scores_gemma":[0.002873174,0.0002821151,0.410477,0.0004914638,0.0001004163,0.00001435515,0.003135724,0.1434285,0.0001287063,0.06773034,0.3687006,0.002637703],"study_design_candidate":"observational","study_design_consensus":"observational","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9655433,0.0009732451,0.0001506239,0.0008212372,0.0002751205,0.0002697629,0.000809275,0.00007200372,0.03108548],"genre_scores_gemma":[0.9908274,0.0001332841,0.006212247,0.001935397,0.0001816685,0.00001172906,0.0001952944,0.00002885175,0.0004740684],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.3677624,"threshold_uncertainty_score":0.9999719,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.0224868405747611,"score_gpt":0.2376843678541349,"score_spread":0.2151975272793737,"validation_status":"score_only:v0-immature-baseline","note":"Baseline scores from an immature model (maturity gate not passed). Scores rank; they never assert a category."}}