{"id":"W4407330655","doi":"10.2139/ssrn.5128896","title":"Enhancing ESG News Annotation: Leveraging GPT for the Analysis of ESG News and Events","year":2025,"lang":"en","type":"preprint","venue":"SSRN Electronic Journal","topic":"Computational and Text Analysis Methods","field":"Social Sciences","cited_by":0,"is_retracted":false,"has_abstract":false,"ca_institutions":"Université de Sherbrooke","funders":"","keywords":"Business; News analytics; Annotation; News media; Advertising; Computer science; Finance","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":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.00424513,0.0001956676,0.0005581254,0.0005521381,0.0009207161,0.0001400196,0.0005084554,0.0001408539,0.00002304602],"category_scores_gemma":[0.0005409414,0.000155053,0.0006764153,0.001089422,0.00008647837,0.0001080088,0.0001634594,0.001244836,3.723883e-7],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0006242694,"about_ca_system_score_gemma":0.004489137,"about_ca_topic_candidate":true,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00431159,"about_ca_topic_score_gemma":0.04513249,"domain_scores_codex":[0.9969992,0.0005863296,0.0006283759,0.0003438805,0.0004836539,0.0009585758],"domain_scores_gemma":[0.9972332,0.001372672,0.0006986686,0.000193519,0.0004299442,0.00007205296],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"theoretical_or_conceptual","study_design_scores_codex":[0.0000934224,0.00008570568,0.02692115,0.00007037518,0.01446106,0.000001796071,0.01388194,0.05649065,0.00001710686,0.1259713,0.0001043018,0.7619013],"study_design_scores_gemma":[0.0005234771,0.00005333071,0.009602197,0.0001301629,0.007653124,0.000007699139,0.0227788,0.02003926,0.00002155297,0.9342474,0.004551109,0.000391845],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.05092319,0.007675715,0.9346654,0.005169648,0.0004935389,0.0003585784,0.000008789401,0.00001834629,0.0006868087],"genre_scores_gemma":[0.9815894,0.009899911,0.00289391,0.0001662189,0.0006848784,0.00002825449,0.00002063627,0.00001151872,0.00470523],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.9317715,"threshold_uncertainty_score":0.9722914,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03073698495373214,"score_gpt":0.3755013808204492,"score_spread":0.3447643958667171,"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."}}