{"id":"W4251852459","doi":"10.1007/978-1-4939-7131-2_101458","title":"Web Crawling","year":2018,"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; World Wide Web; Web application; Computer science; Biology; 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":["insufficient_payload"],"category_scores_codex":[0.0001507609,0.0001732063,0.0002121863,0.0001606745,0.00007239405,0.0001953953,0.001104599,0.0001375756,0.002315418],"category_scores_gemma":[0.00000993144,0.0001448417,0.0001321489,0.00003674484,0.00004302216,0.0001630086,0.0004363117,0.00013971,0.004880175],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00001618384,"about_ca_system_score_gemma":0.00007760266,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.000006263807,"about_ca_topic_score_gemma":0.00001525855,"domain_scores_codex":[0.9989769,0.000004713906,0.0001787162,0.0004571599,0.0002295009,0.0001530083],"domain_scores_gemma":[0.9987636,0.00003361774,0.00008995183,0.0009719732,0.000069285,0.00007153073],"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":[6.872123e-7,0.0000076267,0.00000501228,0.00001390569,0.0001660446,0.00004741163,0.0001208215,0.000001065786,0.0000738844,0.7312343,0.2220347,0.04629452],"study_design_scores_gemma":[0.00004662451,0.00002369664,7.306136e-7,0.00005655103,0.00002596423,0.000006527514,0.000001830531,0.008769427,0.0000621411,0.01704737,0.973701,0.0002581342],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"other","genre_gemma":"other","genre_scores_codex":[0.000003292452,0.0000783007,0.1397519,0.0002204215,0.0002101817,0.00001894657,0.000007090384,0.0002147947,0.8594951],"genre_scores_gemma":[0.0004789628,0.00004409482,0.03635012,0.0003660329,0.0003700363,6.902611e-7,0.00001690482,0.00001442859,0.9623587],"genre_candidate":"other","genre_consensus":"other","teacher_disagreement_score":0.7516663,"threshold_uncertainty_score":0.9985966,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02558767028100625,"score_gpt":0.2337072557640603,"score_spread":0.2081195854830541,"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."}}