{"id":"W4230874668","doi":"10.1007/978-1-4614-6170-8_100858","title":"Web Crawling","year":2014,"lang":"en","type":"book-chapter","venue":"","topic":"Web Data Mining and Analysis","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":false,"ca_institutions":"University of Calgary","funders":"","keywords":"Crawling; Computer science; World Wide Web; Medicine; Anatomy","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":["insufficient_payload"],"consensus_categories":[],"category_scores_codex":[0.000185642,0.0001790689,0.0002527277,0.0001603171,0.00006345143,0.000184252,0.001077395,0.0001371184,0.000449387],"category_scores_gemma":[0.00001198084,0.0001504654,0.0001484885,0.00002944614,0.00002321084,0.00008742177,0.0003567878,0.0001759933,0.002213693],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0000138868,"about_ca_system_score_gemma":0.00005491516,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.000007738409,"about_ca_topic_score_gemma":0.0000113127,"domain_scores_codex":[0.9989806,0.000006739768,0.0001889501,0.000441709,0.0002320065,0.0001499296],"domain_scores_gemma":[0.9987426,0.00005553768,0.00009773382,0.0009836113,0.00004487563,0.00007566417],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"not_applicable","study_design_scores_codex":[1.511983e-7,0.000001487624,0.000002159125,0.000006590889,0.0000422497,0.000009753047,0.000012506,0.000003215998,0.00002075503,0.9182426,0.02360931,0.05804926],"study_design_scores_gemma":[0.00004915471,0.00001427974,8.518891e-7,0.00004952863,0.00002414556,0.000005186322,5.584887e-7,0.02139761,0.00002739536,0.009136785,0.9690472,0.0002472497],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"other","genre_gemma":"other","genre_scores_codex":[6.048739e-7,0.0000540618,0.371348,0.0002583654,0.0001220143,0.00001310162,0.00000286036,0.000161093,0.62804],"genre_scores_gemma":[0.00177506,0.00004147185,0.02737871,0.0005312638,0.0002357245,8.082594e-7,0.00001851845,0.00001538174,0.9700031],"genre_candidate":"other","genre_consensus":"other","teacher_disagreement_score":0.945438,"threshold_uncertainty_score":0.9985632,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01957269536115172,"score_gpt":0.2193976681005148,"score_spread":0.1998249727393631,"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."}}