{"id":"W6944136180","doi":"10.17632/d7vzpcdzcw.4","title":"CrimeDataBD: A Standard Dataset of Crime Records in Bangladesh for Machine Learning Tasks","year":2025,"lang":"en","type":"dataset","venue":"Mendeley Data","topic":"","field":"","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Université de Sherbrooke","funders":"","keywords":"Data source; Crime analysis; Socioeconomic status; Crime scene; Random forest; Distribution (mathematics)","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":"codex-gemma-dda1882f352a","candidate_categories":["metaepi_narrow","open_science","research_integrity","insufficient_payload"],"consensus_categories":[],"category_scores_codex":[0.005788364,0.001213292,0.002354453,0.001765818,0.0002031466,0.0001856009,0.008474605,0.0007339993,0.001598705],"category_scores_gemma":[0.006036872,0.001272604,0.0001539349,0.001308129,0.0002481924,0.001056087,0.00750915,0.002328853,0.0002015036],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0005373491,"about_ca_system_score_gemma":0.001565567,"about_ca_topic_candidate":true,"about_ca_topic_consensus":true,"about_ca_topic_score_codex":0.01285509,"about_ca_topic_score_gemma":0.0179784,"domain_scores_codex":[0.9913012,0.0008542707,0.002351489,0.00280568,0.001332625,0.001354689],"domain_scores_gemma":[0.9857698,0.001128584,0.001461477,0.01109688,0.0002770128,0.0002662397],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"not_applicable","study_design_gemma":"not_applicable","study_design_scores_codex":[0.00169993,0.0003739109,0.00004397739,0.002316789,0.0004139681,0.00008092979,0.00001546571,0.00001431594,0.00007033213,0.0000146738,0.9930579,0.001897785],"study_design_scores_gemma":[0.003456018,0.0003794374,0.000007974184,0.001190614,0.0009448412,0.00001541088,0.00006375782,0.001746517,0.0001198492,0.00005623213,0.9909611,0.00105822],"study_design_candidate":"not_applicable","study_design_consensus":"not_applicable","genre_codex":"dataset","genre_gemma":"dataset","genre_scores_codex":[0.000003031841,0.002358857,0.0004396947,0.00008916765,0.0006986489,0.00204916,0.9941631,0.0001149788,0.00008336313],"genre_scores_gemma":[0.000003393926,0.001196486,0.007860403,0.0002218039,0.0002641971,0.0002455463,0.989816,0.0001667552,0.0002254693],"genre_candidate":"dataset","genre_consensus":"dataset","teacher_disagreement_score":0.008291202,"threshold_uncertainty_score":0.9999728,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.06910668295656398,"score_gpt":0.3673507676158333,"score_spread":0.2982440846592693,"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."}}