{"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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0009894794,0.001651602,0.001125061,0.008280111,0.001435862,0.00365531,0.001630475,0.001250461,0.01616872],"category_scores_gemma":[0.004135352,0.0008892734,0.001058237,0.008596869,0.0006370346,0.004728739,0.002262763,0.001679888,0.02819939],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0008890146,"about_ca_system_score_gemma":0.001453658,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002141871,"about_ca_topic_score_gemma":0.004573595,"domain_scores_codex":[0.9984474,0.0001455422,0.0001114173,0.0003450811,0.0008565951,0.00009407708],"domain_scores_gemma":[0.9977279,0.000472905,0.0001157986,0.000873209,0.0007109298,0.00009920331],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"not_applicable","study_design_scores_codex":[0.00001969843,0.00008146495,0.0004522261,0.0004793928,0.00003531504,0.00007976351,0.0001181941,0.0009528638,0.003203551,0.01192346,0.1754078,0.8072462],"study_design_scores_gemma":[0.00001359288,0.00004711531,0.00183972,0.0004905082,0.00008398266,0.00157828,0.0001200882,0.01927163,0.01634748,0.04386087,0.9162918,0.00005491448],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.01600075,0.03072322,0.5129437,0.005087267,0.003184715,0.0008588398,0.00829218,0.05282127,0.3700881],"genre_scores_gemma":[0.0574109,0.02943231,0.428657,0.002380875,0.001866885,0.0005136227,0.02767085,0.01064541,0.4414222],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.01616872,"threshold_uncertainty_score":0.05408973,"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."}}