{"id":"W2147608180","doi":"10.1109/hicss.2009.68","title":"An N-Gram Based Approach to Automatically Identifying Web Page Genre","year":2009,"lang":"en","type":"article","venue":"","topic":"Authorship Attribution and Profiling","field":"Computer Science","cited_by":27,"is_retracted":false,"has_abstract":true,"ca_institutions":"Dalhousie University","funders":"Killam Trusts","keywords":"Computer science; Web page; Feature (linguistics); Information retrieval; Set (abstract data type); Front page; Listing (finance); World Wide Web; Static web page; Artificial intelligence; Web navigation","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":[0.001501497,0.001303614,0.00104254,0.008518155,0.001127335,0.001435326,0.000848992,0.001146792,0.002718869],"category_scores_gemma":[0.007037117,0.0003021383,0.0009176338,0.005598708,0.0004943883,0.002561122,0.001032006,0.001198802,0.003398503],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0005581747,"about_ca_system_score_gemma":0.0008347665,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0044045,"about_ca_topic_score_gemma":0.009400314,"domain_scores_codex":[0.9982457,0.0005237036,0.0001679891,0.000425091,0.0005061958,0.0001312971],"domain_scores_gemma":[0.9955992,0.002024443,0.0004629182,0.0006380239,0.001027113,0.0002483632],"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":[0.0007351995,0.0006288423,0.01875754,0.0003930122,0.0001660685,0.0002555299,0.0006047529,0.002510432,0.04827377,0.002189423,0.01205384,0.9134315],"study_design_scores_gemma":[0.0001910519,0.0008420915,0.0850508,0.0002124754,0.0003488096,0.002917798,0.001749737,0.7830234,0.05692299,0.0215787,0.04686559,0.0002966226],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.2554587,0.003257402,0.6965175,0.001048669,0.0005485586,0.001320943,0.007385842,0.01866552,0.01579694],"genre_scores_gemma":[0.3553881,0.0007311791,0.6274088,0.0002309513,0.0003086697,0.0004866339,0.008215684,0.0003920235,0.006837876],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.008518155,"threshold_uncertainty_score":0.00909549,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.04993646941629454,"score_gpt":0.3202699916950935,"score_spread":0.270333522278799,"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."}}