{"id":"W3200529564","doi":"10.2139/ssrn.3925176","title":"Measuring information in analyst reports: A machine learning approach","year":2021,"lang":"en","type":"article","venue":"SSRN Electronic Journal","topic":"Auditing, Earnings Management, Governance","field":"Business, Management and Accounting","cited_by":2,"is_retracted":false,"has_abstract":false,"ca_institutions":"University of Calgary; University of Toronto","funders":"","keywords":"Measure (data warehouse); Content (measure theory); Natural language processing; Computer science; Information retrieval; Artificial intelligence; Content analysis; Data science; Psychology; Data mining; Mathematics; Sociology","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":[],"consensus_categories":[],"category_scores_codex":[0.004267744,0.0008475,0.00123568,0.008913031,0.0005044482,0.003316728,0.001116871,0.001401171,0.0007414218],"category_scores_gemma":[0.0253145,0.000383629,0.0009499812,0.006588826,0.0005091664,0.002576687,0.0007899544,0.001409133,0.0004747293],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0008210134,"about_ca_system_score_gemma":0.0009651415,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002969129,"about_ca_topic_score_gemma":0.002342421,"domain_scores_codex":[0.9963792,0.001368275,0.0005370192,0.0006599964,0.0008885395,0.0001670237],"domain_scores_gemma":[0.9643065,0.02827821,0.003208719,0.001518006,0.002422751,0.0002658683],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.000388818,0.00133356,0.1497842,0.0003123494,0.0006774833,0.0001820114,0.0002493701,0.09725841,0.006950372,0.004605459,0.002254304,0.7360037],"study_design_scores_gemma":[0.00002228685,0.0003437464,0.04453265,0.00006788741,0.0001984722,0.0001393011,0.0001389834,0.9372467,0.00672785,0.009333909,0.001186137,0.00006214869],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.3422913,0.003333887,0.6445481,0.001155028,0.0001514307,0.0002944935,0.002410057,0.001926745,0.003889063],"genre_scores_gemma":[0.8808026,0.0007047397,0.1159412,0.00009003426,0.0002452731,0.00008550131,0.001256973,0.00002912511,0.0008445528],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.008913031,"threshold_uncertainty_score":0.02257019,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.008507482336994011,"score_gpt":0.1886230310022022,"score_spread":0.1801155486652082,"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."}}