{"id":"W4390956213","doi":"10.2139/ssrn.4197489","title":"Measuring Corporate Human Capital Disclosures: Lexicon, Data, Code, and Research Opportunities","year":2024,"lang":"en","type":"article","venue":"SSRN Electronic Journal","topic":"Intellectual Capital and Performance Analysis","field":"Business, Management and Accounting","cited_by":4,"is_retracted":false,"has_abstract":false,"ca_institutions":"University of Waterloo","funders":"","keywords":"Lexicon; Accounting; Business; Code (set theory); Human capital; Computer science; Data science; Natural language processing; Economics; Programming language","routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":["bibliometrics"],"consensus_categories":[],"category_scores_codex":[0.01428328,0.0004053734,0.0005841817,0.02424764,0.001210647,0.01465128,0.000907993,0.001278943,0.002946455],"category_scores_gemma":[0.1270919,0.0003408276,0.0003250139,0.0192879,0.002505251,0.009234287,0.00270771,0.001112416,0.001162854],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.003098013,"about_ca_system_score_gemma":0.003817595,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.003969526,"about_ca_topic_score_gemma":0.004637534,"domain_scores_codex":[0.9848186,0.006869078,0.002209877,0.000855619,0.004820916,0.0004259039],"domain_scores_gemma":[0.8540052,0.1030989,0.01641395,0.007819113,0.01747578,0.001187023],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"observational","study_design_scores_codex":[0.0004293692,0.0002474656,0.3004852,0.001006555,0.0001393383,0.0002491467,0.00628385,0.002376141,0.004003858,0.1411557,0.02452618,0.5190973],"study_design_scores_gemma":[0.0002001847,0.0003757196,0.3876919,0.003748821,0.0004761318,0.001745179,0.01831113,0.04921016,0.02329663,0.3129244,0.2016001,0.0004196602],"study_design_candidate":"observational","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.7335728,0.009674861,0.0665883,0.01302195,0.0005086923,0.0005320631,0.01609162,0.001171484,0.1588382],"genre_scores_gemma":[0.9634882,0.002059709,0.02613547,0.0005676426,0.0002995187,0.0001397547,0.004844177,0.0001707031,0.002294838],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.9757524,"threshold_uncertainty_score":0.07553816,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.2263162557439197,"score_gpt":0.3208079945844212,"score_spread":0.09449173884050152,"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."}}