{"id":"W6912982003","doi":"10.5683/sp3/6ya1fk","title":"Extracting Microdata Using IDLS","year":2000,"lang":"en","type":"dataset","venue":"Borealis","topic":"","field":"","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Alberta","funders":"","keywords":"Microdata (statistics); The Internet; Service (business); Presentation (obstetrics); Data extraction; Mixture model","routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":true,"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":[0.00377228,0.001827576,0.001174588,0.0129504,0.001528904,0.00378708,0.002454913,0.0008314311,0.0158581],"category_scores_gemma":[0.01807625,0.001224742,0.001754206,0.01382229,0.0006854603,0.00314593,0.003879588,0.002026457,0.02957009],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00223516,"about_ca_system_score_gemma":0.005459699,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.02408678,"about_ca_topic_score_gemma":0.03561389,"domain_scores_codex":[0.9951352,0.0007330835,0.0008874364,0.001021147,0.001912243,0.0003109265],"domain_scores_gemma":[0.989971,0.001656198,0.0005665534,0.004887329,0.002470442,0.0004484451],"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":[0.0002092925,0.0001112351,0.005106746,0.001114239,0.0001188439,0.0001573764,0.0002438565,0.001710509,0.003104689,0.006483126,0.8927959,0.08884431],"study_design_scores_gemma":[0.00008486045,0.00003438135,0.006466078,0.0001443218,0.00005372701,0.0001714158,0.000233177,0.005881215,0.007114731,0.007850042,0.9718988,0.0000670879],"study_design_candidate":"not_applicable","study_design_consensus":"not_applicable","genre_codex":"dataset","genre_gemma":"dataset","genre_scores_codex":[0.002047599,0.0002718561,0.02880299,0.0003876114,0.0001954061,0.0004071819,0.9331251,0.02794147,0.006820871],"genre_scores_gemma":[0.0024433,0.0001495077,0.02472971,0.00008231515,0.00002311752,0.0004153183,0.9699522,0.0008451092,0.001359413],"genre_candidate":"dataset","genre_consensus":"dataset","teacher_disagreement_score":0.02408678,"threshold_uncertainty_score":0.05305058,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.0591485156832704,"score_gpt":0.3319703943866871,"score_spread":0.2728218787034167,"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."}}