{"id":"W2952264314","doi":"10.48550/arxiv.0906.4019","title":"On the prevalence and scientific impact of duplicate publications in different scientific fields (1980-2007)","year":2009,"lang":"en","type":"preprint","venue":"ArXiv.org","topic":"Academic integrity and plagiarism","field":"Social Sciences","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Université du Québec à Montréal","funders":"","keywords":"Metadata; Library science; Field (mathematics); Scientific literature; Scientific field; Data science; Information retrieval; Geography; Computer science; World Wide Web; Mathematics; Biology; 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":"metacan-v3-hybrid-931329e0061c","candidate_categories":["metaresearch","bibliometrics","research_integrity"],"consensus_categories":[],"category_scores_codex":[0.0376256,0.0003318365,0.00135115,0.04333766,0.001620717,0.003377903,0.001507279,0.001110879,0.001522723],"category_scores_gemma":[0.1901627,0.0002920138,0.001057248,0.03852438,0.001988918,0.003559541,0.00440543,0.000950937,0.000559721],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.002311776,"about_ca_system_score_gemma":0.002601819,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.003825349,"about_ca_topic_score_gemma":0.002905788,"domain_scores_codex":[0.9623334,0.006654677,0.008222606,0.003863308,0.01756416,0.00136182],"domain_scores_gemma":[0.5733684,0.2218593,0.1181939,0.02313618,0.05745578,0.005986507],"domain_codex":null,"domain_gemma":"evaluation","domain_candidate":"evaluation","domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"observational","study_design_scores_codex":[0.0003663491,0.00005003596,0.9308718,0.001206657,0.0006211278,0.0005478079,0.005155636,0.0004728192,0.000776288,0.001320101,0.003106872,0.05550454],"study_design_scores_gemma":[0.00001438834,0.00007480268,0.9855798,0.0002493006,0.0002466821,0.001647212,0.002155238,0.0003344705,0.001105579,0.001066068,0.007483697,0.00004277418],"study_design_candidate":"observational","study_design_consensus":"observational","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9613622,0.01972376,0.001281332,0.001442255,0.00015537,0.00006646445,0.009692339,0.00008232796,0.00619386],"genre_scores_gemma":[0.9856774,0.004897033,0.001329318,0.0001597735,0.0002408804,0.00006455505,0.006519733,0.00003316532,0.00107817],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.9988891,"threshold_uncertainty_score":0.1989856,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.07229518415060894,"score_gpt":0.3446323455733842,"score_spread":0.2723371614227753,"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."}}