{"id":"W1564399566","doi":"","title":"Accuracy of Approximate String Joins Using Grams","year":2007,"lang":"en","type":"article","venue":"","topic":"Data Quality and Management","field":"Decision Sciences","cited_by":16,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Toronto","funders":"","keywords":"Substring; Joins; Similarity (geometry); Computer science; Data mining; String (physics); String metric; Matching (statistics); Join (topology); String searching algorithm; Algorithm; Pattern matching; Pattern recognition (psychology); Artificial intelligence; Mathematics; Statistics; Data structure","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":[0.02033809,0.000857349,0.002720381,0.004551639,0.001235016,0.005045484,0.001919785,0.00211381,0.001214629],"category_scores_gemma":[0.1085326,0.000538504,0.0009983815,0.007684847,0.001196637,0.006895987,0.002849599,0.001058983,0.000971602],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001052913,"about_ca_system_score_gemma":0.001335187,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002269344,"about_ca_topic_score_gemma":0.00170065,"domain_scores_codex":[0.9678589,0.008517425,0.004050728,0.003383955,0.01517036,0.001018709],"domain_scores_gemma":[0.8588359,0.09468088,0.01169959,0.02441604,0.009555055,0.0008124937],"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":[0.005727037,0.0003259658,0.1251081,0.001079474,0.0008451265,0.0005174399,0.001633654,0.3117005,0.02409271,0.02446376,0.009306186,0.4952001],"study_design_scores_gemma":[0.00008731698,0.0004771537,0.02692482,0.0001157731,0.0001432063,0.001121576,0.0007198964,0.894262,0.03709259,0.03312815,0.005791245,0.0001362198],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"methods","genre_scores_codex":[0.5395008,0.005858895,0.4390107,0.0009595884,0.0002651621,0.0001392484,0.003145372,0.005278416,0.005841808],"genre_scores_gemma":[0.8639737,0.000960259,0.1291524,0.0001126444,0.0001103606,0.00006081359,0.004323641,0.0002720988,0.001033904],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.02033809,"threshold_uncertainty_score":0.1075594,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.3939898281848053,"score_gpt":0.4885599141472641,"score_spread":0.0945700859624588,"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."}}