{"id":"W4318148204","doi":"10.1109/bigdata55660.2022.10020871","title":"Predicting Deduplication Performance: An Analytical Model and Empirical Evaluation","year":2022,"lang":"en","type":"article","venue":"2022 IEEE International Conference on Big Data (Big Data)","topic":"Advanced Data Storage Technologies","field":"Computer Science","cited_by":3,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Alberta","funders":"","keywords":"Data deduplication; Computer science; Data mining; Software versioning; Kernel (algebra); Database; Software; Operating system","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":["open_science"],"consensus_categories":["open_science"],"category_scores_codex":[0.001701643,0.0002260543,0.0001927063,0.0003354184,0.0004392889,0.0003348427,0.00850097,0.00007319442,0.00006331935],"category_scores_gemma":[0.0005306479,0.0002375211,0.00001628743,0.0004312034,0.0001412007,0.003315632,0.008200962,0.0006094473,0.00002077238],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0002641534,"about_ca_system_score_gemma":0.0004423017,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00002880252,"about_ca_topic_score_gemma":0.00006522831,"domain_scores_codex":[0.995834,0.0001565052,0.0004190122,0.00159077,0.001700327,0.000299365],"domain_scores_gemma":[0.9949324,0.0001099511,0.0002466704,0.004366409,0.0002314242,0.0001131842],"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":[0.0000743058,0.0002797578,0.002411202,0.00001040052,0.00005540471,0.00001106999,0.0001712653,0.009577411,0.0008338721,0.02291452,0.009909265,0.9537515],"study_design_scores_gemma":[0.000391696,0.0001534506,0.0013,0.00001283484,0.00002237058,0.00004157196,0.0001525655,0.9886321,0.00008094605,0.004092254,0.004870388,0.0002498038],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.2489682,0.00006396672,0.7290938,0.005338928,0.002642121,0.0006339412,0.01083588,0.0005967025,0.001826397],"genre_scores_gemma":[0.9688495,0.0001761601,0.01224546,0.0005520443,0.0002654276,0.0001213418,0.01771075,0.00001650101,0.00006279691],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.9790547,"threshold_uncertainty_score":0.9998205,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.5054255254777182,"score_gpt":0.4297697773039301,"score_spread":0.07565574817378812,"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."}}