{"id":"W2982683456","doi":"10.1186/s13643-019-1222-2","title":"Performance and usability of machine learning for screening in systematic reviews: a comparative evaluation of three tools","year":2019,"lang":"en","type":"article","venue":"Systematic Reviews","topic":"Meta-analysis and systematic reviews","field":"Decision Sciences","cited_by":132,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Alberta","funders":"Agency for Healthcare Research and Quality; U.S. Department of Health and Human Services","keywords":"Medicine; Usability; Medical physics; MEDLINE; Medical education; Artificial intelligence; Human–computer interaction","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":["metaresearch","metaepi_broad"],"consensus_categories":["metaresearch"],"category_scores_codex":[0.5535547,0.000525967,0.02087467,0.0004805022,0.00007635887,0.0002006974,0.001215172,0.000104241,0.0007125232],"category_scores_gemma":[0.1958995,0.0002277845,0.002036048,0.001581412,0.00007587133,0.0005511816,0.0001271805,0.0001849362,0.0004266822],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00006701011,"about_ca_system_score_gemma":0.00006739065,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00001563102,"about_ca_topic_score_gemma":0.00007737125,"domain_scores_codex":[0.8387159,0.09812988,0.05302699,0.001634728,0.008017679,0.0004748144],"domain_scores_gemma":[0.9421263,0.02160151,0.0298222,0.00391149,0.002436078,0.000102418],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"systematic_review","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.00001108439,0.00005224637,0.07527244,0.9177338,0.0001875743,6.095885e-8,0.001900623,0.0006682497,0.00008566993,0.0002962151,0.0001539862,0.003637978],"study_design_scores_gemma":[0.0008484511,0.0002128641,0.008648719,0.1859221,0.001873346,0.00001315402,0.001501592,0.7990519,0.00002229775,0.0005132137,0.001048635,0.0003437189],"study_design_candidate":"systematic_review","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.7170422,0.1172705,0.02563772,0.00004393949,0.0002726045,0.1369427,0.00001851827,0.000006812387,0.002764942],"genre_scores_gemma":[0.9885702,0.0005889155,0.006096224,0.00001583774,0.00001573331,0.004151847,0.000009322037,0.00001434136,0.0005376093],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.7983837,"threshold_uncertainty_score":0.9936615,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.8283611526367183,"score_gpt":0.548349083404606,"score_spread":0.2800120692321123,"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."}}