{"id":"W2038522479","doi":"10.5555/3191835.3191975","title":"Estimating the size of hidden data sources by queries","year":2014,"lang":"en","type":"article","venue":"","topic":"Web Data Mining and Analysis","field":"Computer Science","cited_by":1,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Windsor","funders":"","keywords":"Computer science; Variance (accounting); Variety (cybernetics); Data mining; Sample (material); Competitor analysis; Sample size determination; Baseline (sea); Information retrieval; Artificial intelligence; Statistics; Mathematics","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":[],"consensus_categories":[],"category_scores_codex":[0.0005284902,0.00004991339,0.00009266746,0.00001178282,0.00009035754,0.0001217865,0.002135007,0.00001267061,0.00002515959],"category_scores_gemma":[0.0004170258,0.00002831528,0.00001738037,0.0001510964,0.00005751189,0.0003394569,0.000799695,0.0000373359,0.00001475918],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.000001385439,"about_ca_system_score_gemma":0.00001042061,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0002303418,"about_ca_topic_score_gemma":0.00001677662,"domain_scores_codex":[0.9993877,0.00005217125,0.0001269899,0.0001954858,0.0001456376,0.00009196538],"domain_scores_gemma":[0.998185,0.0004090967,0.00007295731,0.00129545,0.00001653666,0.00002094834],"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":[0.00000279626,0.00009890489,0.01748706,0.00005593678,0.0001878318,0.000001206552,0.003530918,0.0004494838,0.004763859,0.08436017,0.1999486,0.6891133],"study_design_scores_gemma":[0.00005857884,0.00001927879,0.0006032708,0.00001297348,0.00001495299,0.000001708493,0.0001081818,0.9814032,0.001677644,0.001034919,0.01497946,0.00008584763],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.0507069,0.0000528183,0.9445424,0.001946972,0.00005338907,0.00001691485,0.00001385638,0.00006943389,0.002597314],"genre_scores_gemma":[0.7212636,9.892923e-7,0.2779098,0.0002256717,0.00003859753,7.338006e-7,0.00001138739,0.00000219201,0.0005470266],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.9809537,"threshold_uncertainty_score":0.3967411,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.0210675277415002,"score_gpt":0.2535414358595014,"score_spread":0.2324739081180012,"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."}}