{"id":"W2953029151","doi":"10.1186/s13643-019-1062-0","title":"A question of trust: can we build an evidence base to gain trust in systematic review automation technologies?","year":2019,"lang":"en","type":"article","venue":"Systematic Reviews","topic":"Meta-analysis and systematic reviews","field":"Decision Sciences","cited_by":162,"is_retracted":false,"has_abstract":true,"ca_institutions":"Ottawa Hospital","funders":"National Institute of Environmental Health Sciences; Medical Research Council; University College London; U.S. Environmental Protection Agency","keywords":"Automation; Systematic review; Workflow; Documentation; Computer science; Disk formatting; Data science; Medicine; Knowledge management; Risk analysis (engineering); Software engineering; MEDLINE; Engineering","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_narrow","metaepi_broad","insufficient_payload"],"consensus_categories":["metaresearch","insufficient_payload"],"category_scores_codex":[0.3559934,0.0009114734,0.02261245,0.001536169,0.00008075865,0.0005114553,0.004610179,0.0002556699,0.001461051],"category_scores_gemma":[0.3252701,0.0004154204,0.00252471,0.006303442,0.00005657635,0.0008922685,0.0002588947,0.0003105793,0.00919258],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.000337457,"about_ca_system_score_gemma":0.0001643203,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0000672545,"about_ca_topic_score_gemma":0.0001533308,"domain_scores_codex":[0.883239,0.06639046,0.04095171,0.001977456,0.006850726,0.0005905951],"domain_scores_gemma":[0.9509626,0.007961066,0.02526214,0.01390059,0.001626731,0.0002868814],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"systematic_review","study_design_gemma":"systematic_review","study_design_scores_codex":[0.000002882638,0.00006889466,0.001500843,0.9924113,0.0000752613,0.000004390344,0.0006936418,0.00004516551,0.0001428042,0.001863542,0.002266371,0.0009249172],"study_design_scores_gemma":[0.0001650323,0.0001876685,0.0001400337,0.9785631,0.001246926,0.00006392883,0.002081013,0.01495278,0.00004328687,0.001312612,0.0007269988,0.0005165663],"study_design_candidate":"systematic_review","study_design_consensus":"systematic_review","genre_codex":"review","genre_gemma":"empirical","genre_scores_codex":[0.04158738,0.6995155,0.02889466,0.007460945,0.001250675,0.2191037,0.0000528854,0.000218996,0.001915182],"genre_scores_gemma":[0.8791717,0.04281039,0.03316802,0.002430567,0.0001076369,0.02790168,0.00003853346,0.0001821403,0.0141893],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.8375844,"threshold_uncertainty_score":0.9998298,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.4757505986788351,"score_gpt":0.4983922702418923,"score_spread":0.02264167156305724,"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."}}