{"id":"W4390145251","doi":"10.18280/isi.280626","title":"Enhancing Identity Document Classification in KYC Processes: An Evaluation of the Bag-of-Visual-Words Model and Segmentation Impact","year":2023,"lang":"en","type":"article","venue":"Ingénierie des systèmes d information","topic":"Authorship Attribution and Profiling","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Binus University","keywords":"Identity (music); Segmentation; Artificial intelligence; Pattern recognition (psychology); Natural language processing; Computer science; Information retrieval; Art; Aesthetics","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":[],"consensus_categories":[],"category_scores_codex":[0.005191813,0.001930373,0.001641913,0.002615997,0.0008630239,0.002434038,0.001972578,0.0023319,0.001430717],"category_scores_gemma":[0.009877728,0.000324382,0.001308104,0.001922696,0.000901235,0.003440057,0.001480915,0.00212886,0.00183456],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.002335644,"about_ca_system_score_gemma":0.002101396,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.02311003,"about_ca_topic_score_gemma":0.01403682,"domain_scores_codex":[0.9973584,0.0006737448,0.0002049821,0.0007552498,0.0007212911,0.0002863217],"domain_scores_gemma":[0.9940959,0.002935456,0.0003853859,0.0009228463,0.001299281,0.000361141],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.004199537,0.002684898,0.02088334,0.0008138539,0.0003727657,0.0002198207,0.000451416,0.1697831,0.01554301,0.001460992,0.01073751,0.7728499],"study_design_scores_gemma":[0.0000776633,0.0007916895,0.00536092,0.00004150438,0.00008135336,0.00008893084,0.0002702577,0.9767945,0.01367819,0.0007047933,0.0020683,0.0000418443],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.8574693,0.004712677,0.1175985,0.001109623,0.0007128658,0.0009005829,0.001943839,0.009454765,0.00609786],"genre_scores_gemma":[0.8318546,0.0009206034,0.1573819,0.0003291523,0.0001314139,0.000220995,0.005736094,0.0002844965,0.003140612],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.02311003,"threshold_uncertainty_score":0.04595101,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.05658156724440002,"score_gpt":0.361020012995824,"score_spread":0.304438445751424,"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."}}