{"id":"W3203297670","doi":"10.2196/33124","title":"Toward Automated Data Extraction According to Tabular Data Structure: Cross-sectional Pilot Survey of the Comparative Clinical Literature","year":2021,"lang":"en","type":"review","venue":"JMIR Formative Research","topic":"Meta-analysis and systematic reviews","field":"Decision Sciences","cited_by":8,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"","keywords":"Data extraction; Categorical variable; Computer science; Context (archaeology); Metadata; Information retrieval; Statistic; Descriptive statistics; Data mining; Statistics; Data science; MEDLINE; Mathematics; Machine learning; World Wide Web; Geography","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":"metacan-v3-hybrid-931329e0061c","candidate_categories":["metaresearch"],"consensus_categories":["metaresearch"],"category_scores_codex":[0.499395,0.001596104,0.003579032,0.04313672,0.00236124,0.009292524,0.003153045,0.001771264,0.005939225],"category_scores_gemma":[0.7572629,0.002272166,0.005696228,0.03867993,0.003834461,0.0129124,0.009338926,0.002324934,0.002363691],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.007178319,"about_ca_system_score_gemma":0.02466066,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.004649885,"about_ca_topic_score_gemma":0.008097264,"domain_scores_codex":[0.4987423,0.242059,0.1962435,0.01700882,0.04320634,0.002739981],"domain_scores_gemma":[0.1014764,0.6080505,0.1225799,0.05025192,0.1157412,0.00190029],"domain_codex":null,"domain_gemma":"methods","domain_candidate":"methods","domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"observational","study_design_scores_codex":[0.002164092,0.0006486811,0.3388958,0.1869831,0.005884009,0.0008847891,0.07687911,0.00189747,0.007849515,0.01266368,0.02993368,0.3353161],"study_design_scores_gemma":[0.00222874,0.004811017,0.464104,0.2159424,0.0104533,0.003006209,0.0725265,0.01598649,0.01134565,0.02569536,0.1727697,0.001130599],"study_design_candidate":"observational","study_design_consensus":"observational","genre_codex":"methods","genre_gemma":"review","genre_scores_codex":[0.3623171,0.03863116,0.3888912,0.01465805,0.001329645,0.128393,0.05247233,0.003138184,0.0101693],"genre_scores_gemma":[0.4042577,0.01049788,0.4265189,0.005909239,0.0003357846,0.1317264,0.01908567,0.0006789079,0.0009894826],"genre_candidate":"review","genre_consensus":null,"teacher_disagreement_score":0.500605,"threshold_uncertainty_score":0.6173351,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.9763549273064586,"score_gpt":0.767177744680884,"score_spread":0.2091771826255746,"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."}}