{"id":"W4246852754","doi":"10.1145/584800.584802","title":"Topic-oriented collaborative crawling","year":2002,"lang":"en","type":"article","venue":"","topic":"Web Data Mining and Analysis","field":"Computer Science","cited_by":9,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Waterloo","funders":"","keywords":"Web crawler; Crawling; Computer science; Focused crawler; Hash function; Information retrieval; Web page; World Wide Web; Context (archaeology); Data mining; Database; Static web page; Web navigation; Computer security","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":[0.004658248,0.0008305669,0.001757508,0.002500096,0.001576896,0.002564414,0.002594139,0.001310078,0.001554433],"category_scores_gemma":[0.01484193,0.0007541631,0.0007686979,0.002926901,0.0009476498,0.003164283,0.002744594,0.0009954751,0.001801964],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0006836327,"about_ca_system_score_gemma":0.001683814,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002256698,"about_ca_topic_score_gemma":0.004499533,"domain_scores_codex":[0.9939374,0.002154133,0.0003981176,0.001153635,0.00204515,0.0003115177],"domain_scores_gemma":[0.9864395,0.00410019,0.0008871852,0.004781169,0.003238152,0.000553832],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0004049704,0.0009002407,0.009167388,0.0008503404,0.0004410791,0.0004592848,0.002655813,0.07173027,0.04587611,0.0383333,0.01899457,0.8101867],"study_design_scores_gemma":[0.0002509749,0.0005369599,0.006021608,0.00008183948,0.0003390273,0.001272394,0.001052231,0.8265933,0.04507635,0.06218198,0.05645362,0.000139629],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.02742234,0.0004618834,0.9633326,0.0001467744,0.00005380499,0.000442356,0.00008592719,0.003614099,0.004440297],"genre_scores_gemma":[0.2737478,0.0005278944,0.7158141,0.0001740301,0.0001625497,0.0008163683,0.0006718443,0.0004846431,0.007600824],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.004658248,"threshold_uncertainty_score":0.02463549,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01527913115665228,"score_gpt":0.2217458562176939,"score_spread":0.2064667250610417,"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."}}