{"id":"W2110780226","doi":"10.1145/1031453.1031458","title":"Probabilistic models for focused web crawling","year":2004,"lang":"en","type":"article","venue":"","topic":"Web Data Mining and Analysis","field":"Computer Science","cited_by":36,"is_retracted":false,"has_abstract":true,"ca_institutions":"Dalhousie University","funders":"Mitacs","keywords":"CRFS; Crawling; Computer science; Conditional random field; Hidden Markov model; Web crawler; Probabilistic logic; Context (archaeology); Relevance (law); Artificial intelligence; Web page; Machine learning; Information retrieval; Data mining; World Wide Web","routes":{"ca_aff":true,"ca_fund":true,"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":[0.003416491,0.0009118593,0.001322994,0.002919164,0.0008394952,0.002074781,0.002630862,0.0023953,0.003583713],"category_scores_gemma":[0.01888295,0.00135622,0.001622349,0.002771269,0.0008107136,0.003769011,0.00110517,0.001801885,0.001767436],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001294737,"about_ca_system_score_gemma":0.001381719,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01110845,"about_ca_topic_score_gemma":0.01399539,"domain_scores_codex":[0.9979795,0.0007983723,0.0001587919,0.0004190495,0.0005387891,0.0001054887],"domain_scores_gemma":[0.9877757,0.009223141,0.0006189485,0.001276392,0.0009460825,0.0001596842],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.00009509427,0.0001132189,0.001915509,0.0001827391,0.0001345593,0.0002233686,0.0002498979,0.8500144,0.001240219,0.06529982,0.006349663,0.0741815],"study_design_scores_gemma":[0.00001160885,0.000008815004,0.0002431749,0.00001000081,0.00001461418,0.00004778799,0.000008449565,0.9641122,0.0001928712,0.03371776,0.001616699,0.00001606944],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.005623993,0.0003830525,0.9906483,0.0001994298,0.00002721714,0.00006482,0.000469889,0.001583442,0.0009999096],"genre_scores_gemma":[0.2536176,0.001748168,0.7343921,0.0002613276,0.0002387991,0.0009720279,0.003065038,0.0006217817,0.0050832],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.01110845,"threshold_uncertainty_score":0.02208757,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03819472388232283,"score_gpt":0.2513862193380534,"score_spread":0.2131914954557306,"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."}}