{"id":"W4414059432","doi":"10.1007/s10032-025-00555-5","title":"Optimizing identity documents classification in online systems: A comparative analysis","year":2025,"lang":"en","type":"article","venue":"International Journal on Document Analysis and Recognition (IJDAR)","topic":"Text and Document Classification Technologies","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":false,"ca_institutions":"International Medias Data Services (Canada)","funders":"","keywords":"Convolutional neural network; Context (archaeology); Identity (music); Process (computing); A priori and a posteriori; Identity management; Strengths and weaknesses","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":[],"consensus_categories":[],"category_scores_codex":[0.003885979,0.0005173509,0.001255485,0.00181442,0.0006531327,0.002582298,0.0009905263,0.0008646972,0.002911661],"category_scores_gemma":[0.01319964,0.0002056212,0.000579562,0.002570336,0.000477263,0.002990885,0.0006528181,0.0005551391,0.0006009049],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.002479901,"about_ca_system_score_gemma":0.001911705,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.008352025,"about_ca_topic_score_gemma":0.006963532,"domain_scores_codex":[0.9962226,0.001597066,0.0002662026,0.0004376067,0.001088946,0.0003874831],"domain_scores_gemma":[0.9834254,0.01220448,0.0005860263,0.001238569,0.002346042,0.0001994497],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"observational","study_design_scores_codex":[0.001800754,0.0009792936,0.01390687,0.0003372811,0.0001661253,0.00006942569,0.0001829324,0.1799626,0.00524734,0.006726855,0.003333638,0.787287],"study_design_scores_gemma":[0.00003199066,0.0005927195,0.00863424,0.00001679932,0.0001404651,0.00009505334,0.0003050436,0.9785832,0.006222963,0.00367755,0.001679467,0.00002047627],"study_design_candidate":"observational","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.7729157,0.009723417,0.2010436,0.0008119297,0.0001102922,0.0002236561,0.0003382258,0.001289427,0.01354372],"genre_scores_gemma":[0.9567384,0.001053563,0.03946256,0.00004475101,0.00005290896,0.00003120682,0.0003175625,0.0001048669,0.002194254],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.008352025,"threshold_uncertainty_score":0.02055132,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.04133032381567437,"score_gpt":0.3529996195214886,"score_spread":0.3116692957058143,"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."}}