{"id":"W2994977308","doi":"10.2196/16606","title":"Use of Artificial Intelligence for Medical Literature Search: Randomized Controlled Trial Using the Hackathon Format","year":2019,"lang":"en","type":"article","venue":"Interactive Journal of Medical Research","topic":"Biomedical and Engineering Education","field":"Engineering","cited_by":38,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"","keywords":"Computer science; Search engine; Information retrieval; Relevance (law); Field (mathematics); Scientific literature; Artificial intelligence; Data science; Mathematics","routes":{"ca_aff":false,"ca_fund":false,"ca_venue":true,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":["metaresearch"],"consensus_categories":[],"category_scores_codex":[0.01719748,0.0001417689,0.000933475,0.0004468281,0.0000511552,0.0001178794,0.0006334318,0.0003214145,0.0006287111],"category_scores_gemma":[0.0258816,0.000074002,0.0004625967,0.0004442281,0.0003530096,0.0003329673,0.00008338305,0.002148374,0.000008312662],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001331954,"about_ca_system_score_gemma":0.0005259941,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00001564091,"about_ca_topic_score_gemma":0.00000271388,"domain_scores_codex":[0.9940234,0.0007316854,0.001282398,0.0001161711,0.00346352,0.0003827879],"domain_scores_gemma":[0.9814593,0.01674866,0.0001629221,0.0001810167,0.001055754,0.000392369],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"randomized_trial","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.9035926,0.0004199669,0.000002679384,0.0006197068,0.0008741547,0.000038241,0.003973937,0.001976632,0.002620987,0.002591941,0.002956029,0.08033311],"study_design_scores_gemma":[0.1780293,0.0004682865,0.000002329933,0.002019924,0.00004787863,0.0001041465,0.000719331,0.8118776,0.002761378,0.001255495,0.002607234,0.0001071075],"study_design_candidate":"randomized_trial","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.7288039,0.001801397,0.2538595,0.006514495,0.006051085,0.002833877,0.00001010002,0.00002218004,0.0001034153],"genre_scores_gemma":[0.996257,0.001125231,0.001003793,0.00005758483,0.00141718,0.00005660237,0.000003743592,0.00002544918,0.00005343165],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.809901,"threshold_uncertainty_score":0.9823238,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1099685439024809,"score_gpt":0.4297201560263877,"score_spread":0.3197516121239068,"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."}}