{"id":"W2096575814","doi":"10.1109/wi.2006.50","title":"Binary Cybergenre Classification Using Theoretic Feature Measures","year":2006,"lang":"en","type":"article","venue":"","topic":"Authorship Attribution and Profiling","field":"Computer Science","cited_by":9,"is_retracted":false,"has_abstract":true,"ca_institutions":"Dalhousie University","funders":"","keywords":"Feature selection; Computer science; Feature (linguistics); Binary classification; Artificial intelligence; Context (archaeology); Set (abstract data type); Pattern recognition (psychology); Selection (genetic algorithm); Binary number; Feature extraction; Web page; Data mining; Machine learning; Support vector machine; Mathematics; 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":[0.00817053,0.0007441569,0.001085949,0.001945411,0.0004720069,0.001958314,0.0007413735,0.0008055169,0.001220594],"category_scores_gemma":[0.0492203,0.0001559675,0.0006974539,0.00125828,0.001107831,0.002768328,0.0009258316,0.0008873804,0.0002750454],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0007505124,"about_ca_system_score_gemma":0.0003810087,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0003629577,"about_ca_topic_score_gemma":0.0003305599,"domain_scores_codex":[0.9948121,0.002335527,0.0005004145,0.0007045838,0.001398614,0.0002487031],"domain_scores_gemma":[0.9425159,0.04695521,0.003218364,0.0042537,0.002614956,0.0004419218],"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":[0.006088423,0.001891375,0.1391573,0.0007715806,0.0004667958,0.0003646917,0.0006473837,0.06244592,0.06069233,0.01013072,0.001720164,0.7156233],"study_design_scores_gemma":[0.0004421093,0.005294066,0.2164262,0.0001263646,0.0003465332,0.001081787,0.0006304237,0.6807295,0.06541751,0.02724778,0.002012284,0.0002454986],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9168207,0.000381986,0.08040264,0.0002231971,0.0000712435,0.0001495877,0.0002073186,0.0002868685,0.001456319],"genre_scores_gemma":[0.9683544,0.00005025265,0.03097636,0.00004012919,0.00003190949,0.00009334729,0.000254372,0.00001829869,0.000181017],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.00817053,"threshold_uncertainty_score":0.04321039,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.04985563926123843,"score_gpt":0.2836080435254115,"score_spread":0.2337524042641731,"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."}}