{"id":"W2059414344","doi":"10.1007/s10044-014-0432-4","title":"An investigation of implicit features in compression-based learning for comparing webpages","year":2014,"lang":"en","type":"article","venue":"Pattern Analysis and Applications","topic":"Computability, Logic, AI Algorithms","field":"Computer Science","cited_by":2,"is_retracted":false,"has_abstract":false,"ca_institutions":"University of Alberta","funders":"","keywords":"Computer science; Feature (linguistics); Feature selection; Web page; Compression (physics); Task (project management); Phishing; Pattern recognition (psychology); Artificial intelligence; Cover (algebra); Feature vector; Image compression; Space (punctuation); Data compression; Selection (genetic algorithm); Information retrieval; Data mining; Image (mathematics); The Internet; Image processing; 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":[0.00226059,0.000340251,0.0006760525,0.00232556,0.0004833897,0.001881542,0.001142772,0.0009907925,0.002309971],"category_scores_gemma":[0.02328125,0.0002163845,0.0003532133,0.002985837,0.001124654,0.003796689,0.00104948,0.001349466,0.0003765892],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0006349198,"about_ca_system_score_gemma":0.0009220861,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002504683,"about_ca_topic_score_gemma":0.002580544,"domain_scores_codex":[0.9988486,0.0003263947,0.00009837041,0.0002337887,0.0004137996,0.00007910701],"domain_scores_gemma":[0.9785922,0.01600053,0.001208614,0.00180197,0.001997233,0.0003994224],"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":[0.001463043,0.0005695802,0.04648917,0.0004207779,0.0001419208,0.0002466799,0.000518765,0.03274409,0.02775853,0.02188182,0.002290043,0.8654755],"study_design_scores_gemma":[0.00004689413,0.0004882367,0.02850611,0.00005914804,0.00007377905,0.0004503797,0.0002641028,0.9336674,0.01175498,0.02314022,0.00151529,0.00003347545],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.6288707,0.001173482,0.3635167,0.0006446934,0.00008481414,0.0001990003,0.0005276903,0.0007992653,0.004183801],"genre_scores_gemma":[0.9338207,0.0002359633,0.0644173,0.0000512423,0.00006102434,0.00005218125,0.0003937948,0.0000381721,0.0009295033],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.002504683,"threshold_uncertainty_score":0.01195532,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01604752687691857,"score_gpt":0.2793494901573245,"score_spread":0.2633019632804059,"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."}}