{"id":"W2626925008","doi":"10.1145/3106426.3106484","title":"Preference-driven similarity join","year":2017,"lang":"en","type":"article","venue":"Proceedings of the International Conference on Web Intelligence","topic":"Data Quality and Management","field":"Decision Sciences","cited_by":1,"is_retracted":false,"has_abstract":true,"ca_institutions":"Simon Fraser University","funders":"","keywords":"Similarity (geometry); Join (topology); Set (abstract data type); Object (grammar); Variety (cybernetics); Similitude","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":[0.005327923,0.001160352,0.002115099,0.001891453,0.001340373,0.003056451,0.0035425,0.001812085,0.009185323],"category_scores_gemma":[0.01832884,0.0005839521,0.001568236,0.003740165,0.00100761,0.005833348,0.004834525,0.002213614,0.003338663],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0008708166,"about_ca_system_score_gemma":0.002280867,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00187666,"about_ca_topic_score_gemma":0.00234693,"domain_scores_codex":[0.9891009,0.002690785,0.000865886,0.002043511,0.004694311,0.0006045617],"domain_scores_gemma":[0.9902278,0.003226395,0.0006667993,0.002725724,0.002539899,0.0006133456],"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":[0.001697173,0.001088792,0.006630087,0.001169832,0.0005191261,0.0007228518,0.0009925253,0.1590065,0.02605386,0.1411336,0.03195364,0.629032],"study_design_scores_gemma":[0.0001864296,0.0006740185,0.001357779,0.00004923422,0.0001147842,0.001046274,0.0003174527,0.7852316,0.0134692,0.1736632,0.02376205,0.0001279913],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.01515567,0.0004960791,0.9773214,0.0002848669,0.0001350321,0.000382866,0.0005687328,0.001150211,0.004505073],"genre_scores_gemma":[0.3789535,0.0005643865,0.6075401,0.0005318423,0.0003227437,0.0004405221,0.002902014,0.0003976709,0.0083473],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.009185323,"threshold_uncertainty_score":0.03072792,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.4629911135755076,"score_gpt":0.4403104986276323,"score_spread":0.02268061494787527,"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."}}