{"id":"W4409136955","doi":"10.1371/journal.pone.0320151","title":"Using artificial intelligence tools to automate data extraction for living evidence syntheses","year":2025,"lang":"en","type":"article","venue":"PLoS ONE","topic":"Explainable Artificial Intelligence (XAI)","field":"Computer Science","cited_by":9,"is_retracted":false,"has_abstract":true,"ca_institutions":"Simon Fraser University; McMaster University","funders":"Canadian Institutes of Health Research","keywords":"Computer science; Python (programming language); Parsing; Data extraction; Artificial intelligence; Information extraction; Data science; Machine learning; MEDLINE; Programming language","routes":{"ca_aff":true,"ca_fund":true,"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"],"consensus_categories":[],"category_scores_codex":[0.001430357,0.000200716,0.0002916116,0.0003014562,0.000406603,0.0009005539,0.00255095,0.00007191114,0.00003665841],"category_scores_gemma":[0.01994627,0.000221487,0.00005024233,0.001048289,0.00005349683,0.003358648,0.001222142,0.0001532514,0.0001635215],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001799523,"about_ca_system_score_gemma":0.0002428385,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.000180437,"about_ca_topic_score_gemma":0.0001451279,"domain_scores_codex":[0.9974191,0.0001174559,0.0005696623,0.0009189878,0.0004372257,0.0005375971],"domain_scores_gemma":[0.9921721,0.00541138,0.0001442577,0.001792977,0.0003591122,0.0001202392],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"bench_or_experimental","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.00007336675,0.00159787,0.0002623225,0.0004842845,0.0001874318,0.00001699281,0.00110296,0.004284262,0.3759567,0.2471532,0.0002832058,0.3685973],"study_design_scores_gemma":[0.000002787075,0.00005827824,0.00005710548,0.003251239,0.00004796176,0.0000015507,0.0001673329,0.6337525,0.3536388,0.008719475,0.00009519064,0.000207708],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.02772519,0.0002250925,0.9680043,0.002122883,0.0003597303,0.0008643871,0.00001597564,0.0003717077,0.0003107328],"genre_scores_gemma":[0.6666723,0.00004108733,0.3325835,0.0003377185,0.0001359257,0.00007970785,0.000001317575,0.00001363957,0.0001348219],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.6389471,"threshold_uncertainty_score":0.9883091,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.590885932505264,"score_gpt":0.4242192356674022,"score_spread":0.1666666968378618,"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."}}