{"id":"W4236760935","doi":"10.1109/wi.2004.10057","title":"Focused Crawling by Learning HMM from User's Topic-specific Browsing","year":2005,"lang":"en","type":"article","venue":"IEEE/WIC/ACM International Conference on Web Intelligence (WI'04)","topic":"Web Data Mining and Analysis","field":"Computer Science","cited_by":4,"is_retracted":false,"has_abstract":true,"ca_institutions":"Dalhousie University","funders":"","keywords":"Crawling; Computer science; Hidden Markov model; Web page; Information retrieval; Focused crawler; World Wide Web; Web crawler; Session (web analytics); Task (project management); Traverse; Construct (python library); Static web page; Artificial intelligence; Web navigation","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":[0.001158381,0.0009282389,0.001228125,0.002121347,0.0003271926,0.0006997566,0.001102226,0.001129638,0.0006415793],"category_scores_gemma":[0.00491962,0.0007483457,0.0008947196,0.001367338,0.0003341919,0.001819487,0.0005853767,0.000999208,0.0008339487],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0005169235,"about_ca_system_score_gemma":0.0008752326,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.009964434,"about_ca_topic_score_gemma":0.01974324,"domain_scores_codex":[0.9993517,0.0001869117,0.00004361831,0.0002416938,0.000124033,0.00005208116],"domain_scores_gemma":[0.9965718,0.002029615,0.000193154,0.000559684,0.0005112629,0.0001344097],"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":[0.0009174813,0.001287255,0.05852711,0.0004324682,0.0006723835,0.0005942023,0.0009162541,0.2735412,0.02685639,0.00274738,0.01223746,0.6212705],"study_design_scores_gemma":[0.00001564677,0.00009562601,0.00397765,0.000007661928,0.00005230462,0.0001113057,0.00002941005,0.9901406,0.003334442,0.001730135,0.0004857753,0.00001953115],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.2475408,0.0005309444,0.7407102,0.0001377309,0.00003765776,0.0001991028,0.001105029,0.008861635,0.0008769526],"genre_scores_gemma":[0.7068833,0.0003962335,0.2845747,0.0001333336,0.00006246877,0.0003176686,0.00480302,0.0003090965,0.002520294],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.009964434,"threshold_uncertainty_score":0.01981288,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.06551536332156896,"score_gpt":0.3010360112915977,"score_spread":0.2355206479700288,"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."}}