{"id":"W1981802602","doi":"10.1109/wi.2004.70","title":"Focused Crawling by Learning HMM from User's Topic-specific Browsing","year":2004,"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":"Dalhousie University","funders":"","keywords":"Crawling; Computer science; Hidden Markov model; Web page; Focused crawler; Information retrieval; Web crawler; World Wide Web; Session (web analytics); Task (project management); Graph; 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":[0.001223054,0.0009729415,0.001191509,0.002148174,0.0003741027,0.0007159577,0.001089647,0.001235593,0.0005931443],"category_scores_gemma":[0.004921903,0.0007661493,0.0008854956,0.001511572,0.0003875962,0.001905852,0.0005606544,0.001011941,0.000922294],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0005534292,"about_ca_system_score_gemma":0.0008234748,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01031572,"about_ca_topic_score_gemma":0.01972558,"domain_scores_codex":[0.9992613,0.0002547523,0.0000519455,0.000242821,0.0001362314,0.00005286709],"domain_scores_gemma":[0.9967992,0.001788032,0.0002128138,0.0005870112,0.0004965019,0.0001163595],"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":[0.0007704725,0.001082173,0.0600429,0.0004526416,0.0006984618,0.0005180451,0.0007624475,0.2407495,0.02983565,0.00239,0.01110974,0.6515881],"study_design_scores_gemma":[0.00001722897,0.0001098964,0.005710112,0.00001179205,0.00007206888,0.0001524917,0.00003286568,0.9864271,0.005118845,0.001674779,0.0006456372,0.00002709615],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.1889287,0.000576185,0.7994661,0.0001355407,0.00003535315,0.0002045717,0.001178913,0.008665121,0.0008095786],"genre_scores_gemma":[0.6061636,0.0004569361,0.3851881,0.0001182094,0.00006077402,0.0003244304,0.005350731,0.0002410472,0.002096219],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.01031572,"threshold_uncertainty_score":0.02051133,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01297653658873975,"score_gpt":0.2095709369393586,"score_spread":0.1965944003506188,"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."}}