{"id":"W4413822057","doi":"10.1109/tbdata.2025.3604171","title":"Optimizing Deduplication Parameters via a Change-Estimation Analytical Model","year":2025,"lang":"en","type":"article","venue":"IEEE Transactions on Big Data","topic":"Innovation Diffusion and Forecasting","field":"Decision Sciences","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Alberta","funders":"","keywords":"Computer science; Data deduplication; Database","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":[0.001873172,0.001026764,0.0006979965,0.0007834862,0.0005630388,0.001584105,0.001316173,0.000860343,0.001320144],"category_scores_gemma":[0.01254044,0.0005840353,0.0004806084,0.0008961025,0.0006573998,0.002609877,0.0007766869,0.001355783,0.0007188106],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001786041,"about_ca_system_score_gemma":0.002254821,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.005080025,"about_ca_topic_score_gemma":0.005328454,"domain_scores_codex":[0.9987441,0.0002346973,0.0000827096,0.0003523888,0.0004301099,0.0001558675],"domain_scores_gemma":[0.9949856,0.00307235,0.0003665934,0.0005266922,0.0009490116,0.00009972387],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0001550932,0.0001207095,0.002672918,0.0001251972,0.00002677067,0.00008366405,0.0001144812,0.9389396,0.008377264,0.005632885,0.002719905,0.04103151],"study_design_scores_gemma":[0.000004990475,0.00002099625,0.0001365657,0.000005237001,0.000005130988,0.00002406836,0.00001090299,0.9950398,0.003450344,0.0007702663,0.0005264763,0.000005147221],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.1413242,0.001484863,0.8438069,0.001148359,0.0001217273,0.0003640494,0.0007221981,0.004978835,0.006048815],"genre_scores_gemma":[0.8018524,0.0005613284,0.1921322,0.0002895954,0.00003855314,0.0003520839,0.0006965427,0.0003971316,0.003680124],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.005080025,"threshold_uncertainty_score":0.01295865,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.4476223141122868,"score_gpt":0.4268905043208662,"score_spread":0.02073180979142053,"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."}}