{"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":"codex-gemma-dda1882f352a","candidate_categories":["scholarly_communication","open_science","insufficient_payload"],"consensus_categories":[],"category_scores_codex":[0.002055338,0.0001799454,0.0002512807,0.000164045,0.0004061922,0.001391974,0.01010172,0.00007433012,0.001333835],"category_scores_gemma":[0.007898115,0.0001157051,0.0001629923,0.0001386457,0.0004876597,0.0009340014,0.002063349,0.0002898818,0.0003407566],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00005560424,"about_ca_system_score_gemma":0.0001023528,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00008610991,"about_ca_topic_score_gemma":0.00009761292,"domain_scores_codex":[0.9962781,0.00002063821,0.0006961185,0.0005651031,0.002223015,0.0002170234],"domain_scores_gemma":[0.9961894,0.0002095309,0.001119164,0.00079553,0.001604093,0.00008235624],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"theoretical_or_conceptual","study_design_scores_codex":[0.00007340548,0.0001114215,0.01014206,0.00001042363,0.00003718145,5.442178e-7,0.0002339849,0.0000215546,0.004643972,0.9510363,0.01017922,0.02350991],"study_design_scores_gemma":[0.0002257522,0.000161471,0.02808271,0.0004556451,0.00002490556,0.000005072107,0.00211316,0.03226092,0.1036505,0.7953193,0.03730574,0.0003948269],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":"theoretical_or_conceptual","genre_codex":"other","genre_gemma":"empirical","genre_scores_codex":[0.3042265,0.000007593053,0.001024892,0.0316842,0.001937589,0.0004344881,0.0001636173,0.00004581538,0.6604753],"genre_scores_gemma":[0.9913488,0.00008423001,0.0006286705,0.0005037063,0.0000843833,0.00001782373,0.000001574832,0.000006772791,0.00732403],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.6871223,"threshold_uncertainty_score":0.9996447,"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."}}