{"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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.002056832,0.0008941154,0.001287911,0.002259153,0.0005951401,0.001218476,0.002487141,0.00214239,0.00225456],"category_scores_gemma":[0.01077911,0.0009712051,0.001409708,0.002195017,0.0008756345,0.003019386,0.0008960207,0.001522743,0.0009395911],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001233549,"about_ca_system_score_gemma":0.0009343032,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0105761,"about_ca_topic_score_gemma":0.01038833,"domain_scores_codex":[0.9987214,0.0004869833,0.00007896136,0.0003130616,0.0003051279,0.00009443969],"domain_scores_gemma":[0.9931553,0.005360044,0.0004745187,0.0004727246,0.0004541243,0.00008325056],"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":[0.00005230639,0.00005665419,0.001453358,0.00009500521,0.00005947077,0.0001003296,0.000115681,0.9278284,0.0007756218,0.0434447,0.001616058,0.02440235],"study_design_scores_gemma":[0.000005529856,0.000007081818,0.0001698394,0.000004494559,0.000006186133,0.00001794144,0.000003980622,0.9838057,0.00008386149,0.01558251,0.000306296,0.000006574249],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.01945763,0.0005154833,0.9775016,0.0002655941,0.0000255167,0.00006558975,0.0004191735,0.0006881314,0.00106132],"genre_scores_gemma":[0.6437387,0.001644736,0.3440767,0.0002716529,0.0001781951,0.0009232851,0.002506436,0.0002985772,0.006361675],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.0105761,"threshold_uncertainty_score":0.02102906,"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."}}