{"id":"W4398852159","doi":"10.7910/dvn/qtoiqf/jefic1","title":"scrapers_article.RData","year":2019,"lang":"en","type":"dataset","venue":"Harvard Dataverse","topic":"","field":"","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of British Columbia","funders":"","keywords":"Scraper site; Geography; Computer science; World Wide Web","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":["insufficient_payload"],"consensus_categories":[],"category_scores_codex":[0.00135617,0.002365364,0.001748696,0.006115268,0.001139746,0.005099995,0.003482583,0.00256183,0.3540365],"category_scores_gemma":[0.01186629,0.0008843776,0.001587548,0.009648199,0.0006945202,0.002551682,0.003244378,0.001998929,0.4158788],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00188282,"about_ca_system_score_gemma":0.002822368,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.02343009,"about_ca_topic_score_gemma":0.03962404,"domain_scores_codex":[0.9987175,0.0002477849,0.000158726,0.0003580935,0.0002611326,0.0002568373],"domain_scores_gemma":[0.9954757,0.001293609,0.0004154353,0.001163759,0.001160148,0.0004913182],"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":[0.00002207576,0.000005588338,0.00016241,0.0003278095,0.00001121475,0.000005288582,0.00000810379,0.00004570084,0.00002093199,0.0002311345,0.9984183,0.0007413464],"study_design_scores_gemma":[0.0002033304,0.00001384628,0.001284818,0.0003601811,0.00002115995,0.00002507649,0.00006024483,0.0001595178,0.0001806029,0.001285686,0.9963814,0.00002404801],"study_design_candidate":"not_applicable","study_design_consensus":"not_applicable","genre_codex":"dataset","genre_gemma":"dataset","genre_scores_codex":[0.00003156988,0.00004950278,0.00002242721,0.00009271969,0.00003509857,0.000005278242,0.9985982,0.000295429,0.000869789],"genre_scores_gemma":[0.000267135,0.00007720505,0.0001088125,0.0001067245,0.0000239013,0.00004760576,0.9972906,0.0002055013,0.001872603],"genre_candidate":"dataset","genre_consensus":"dataset","teacher_disagreement_score":0.3540365,"threshold_uncertainty_score":0.9213885,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03012375292772027,"score_gpt":0.2667621969458742,"score_spread":0.236638444018154,"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."}}