{"id":"W2135390635","doi":"10.1111/j.1467-8640.2012.00411.x","title":"PROBABILISTIC MODELS FOR FOCUSED WEB CRAWLING","year":2012,"lang":"en","type":"article","venue":"Computational Intelligence","topic":"Web Data Mining and Analysis","field":"Computer Science","cited_by":26,"is_retracted":false,"has_abstract":true,"ca_institutions":"Dalhousie University","funders":"","keywords":"Computer science; Conditional random field; Web crawler; Crawling; Hidden Markov model; Probabilistic logic; Graphical model; Maximum-entropy Markov model; Artificial intelligence; Web page; Markov chain; Statistical model; Data mining; Information retrieval; Machine learning; Markov model; World Wide Web","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":[],"consensus_categories":[],"category_scores_codex":[0.0004348239,0.000118808,0.0001374144,0.0001112186,0.0001591432,0.0001351145,0.0006790468,0.00003713623,0.00001183537],"category_scores_gemma":[0.0001587013,0.0001127798,0.00009084678,0.0003591493,0.00004449924,0.0008481736,0.000137117,0.00007091108,0.00008522197],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0000376939,"about_ca_system_score_gemma":0.00009775555,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.000005084312,"about_ca_topic_score_gemma":0.000001220602,"domain_scores_codex":[0.9988493,0.00003366722,0.0002691339,0.0002942649,0.0002417401,0.000311835],"domain_scores_gemma":[0.998701,0.0005947987,0.00008493027,0.0002824239,0.0002140908,0.0001227934],"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":[0.000002870555,0.00007387056,0.00008823478,0.00001736029,0.00002151058,4.238943e-7,0.0004500988,0.5766867,0.00002647227,0.3873499,0.000418766,0.03486373],"study_design_scores_gemma":[0.00003954427,0.00002017779,0.00004706051,0.00001433025,0.000009103594,0.00000438473,0.00001737201,0.8392422,0.0002413174,0.1596203,0.0006181389,0.0001260982],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.002156929,0.0002211101,0.9963366,0.0003378137,0.0002478654,0.000133901,0.00001532634,0.000128883,0.0004215571],"genre_scores_gemma":[0.6796272,0.00000267891,0.3199988,0.0001425227,0.0001007581,0.00002994771,0.00002243161,0.000005930524,0.00006979528],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.6774703,"threshold_uncertainty_score":0.4599028,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.08854326872503002,"score_gpt":0.3113353877588746,"score_spread":0.2227921190338446,"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."}}