{"id":"W4249958738","doi":"10.1007/978-0-387-39940-9_2347","title":"Data Deduplication","year":2009,"lang":"en","type":"book-chapter","venue":"Encyclopedia of Database Systems","topic":"Data Quality and Management","field":"Decision Sciences","cited_by":0,"is_retracted":false,"has_abstract":false,"ca_institutions":"University of Waterloo","funders":"","keywords":"Data deduplication; Computer science; 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":"codex-gemma-dda1882f352a","candidate_categories":["metaepi_narrow","insufficient_payload"],"consensus_categories":[],"category_scores_codex":[0.006680399,0.0003953569,0.0009025611,0.00051302,0.00009073846,0.0001911999,0.005316386,0.0002318418,0.0007486915],"category_scores_gemma":[0.00161007,0.0003263381,0.0001215527,0.0001625808,0.0001348412,0.00110738,0.002028505,0.0003046765,0.00289609],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00004679003,"about_ca_system_score_gemma":0.0001926147,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0001923554,"about_ca_topic_score_gemma":0.0001616927,"domain_scores_codex":[0.9931735,0.0001651336,0.002110476,0.00147939,0.002797374,0.0002741147],"domain_scores_gemma":[0.9873815,0.0008521845,0.001697429,0.009552355,0.0003354809,0.0001810594],"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":[0.00001049654,0.00003466515,0.000005787407,0.00009350398,0.00004560245,0.00001394161,0.00003017302,0.000006326612,0.000003114163,0.2340324,0.6947393,0.07098471],"study_design_scores_gemma":[0.0001536055,0.00003718665,0.00003072571,0.0002794591,0.0000964646,0.000006539655,0.00009209685,0.0002206151,0.000001281542,0.008305746,0.9904449,0.0003313552],"study_design_candidate":"not_applicable","study_design_consensus":"not_applicable","genre_codex":"other","genre_gemma":"other","genre_scores_codex":[0.000003706767,0.003173708,0.007662242,0.0003004046,0.001545923,0.000871599,0.0337375,0.0000629108,0.952642],"genre_scores_gemma":[0.0001390537,0.003880562,0.001444948,0.0001947229,0.0009278232,0.00001588702,0.0240696,0.00004439173,0.969283],"genre_candidate":"other","genre_consensus":"other","teacher_disagreement_score":0.2957056,"threshold_uncertainty_score":0.9999189,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.2149054609765482,"score_gpt":0.3946475339549821,"score_spread":0.1797420729784339,"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."}}