{"id":"W1996429742","doi":"10.1145/1188966.1189013","title":"Exploring a new space of features for document classification","year":2006,"lang":"en","type":"article","venue":"","topic":"Text and Document Classification Technologies","field":"Computer Science","cited_by":8,"is_retracted":false,"has_abstract":true,"ca_institutions":"Queen's University","funders":"","keywords":"Computer science; Space (punctuation); Information retrieval; Artificial intelligence; Pattern recognition (psychology)","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":[0.002122142,0.0009641879,0.00121532,0.005335013,0.0007171921,0.003036923,0.001259566,0.001074653,0.002014723],"category_scores_gemma":[0.009058546,0.0002804986,0.001312147,0.005454896,0.001042783,0.005466394,0.001351934,0.001325218,0.0008759801],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0007909291,"about_ca_system_score_gemma":0.000762655,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002062953,"about_ca_topic_score_gemma":0.001640538,"domain_scores_codex":[0.9982345,0.0004975963,0.0001964151,0.0004334747,0.0005174,0.0001205003],"domain_scores_gemma":[0.9952765,0.00259747,0.0003813094,0.0006518307,0.0009312838,0.0001616226],"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":[0.0003361673,0.0002962586,0.007041563,0.0004179002,0.0001279792,0.0001935278,0.0003953528,0.0193556,0.01352108,0.02425362,0.007077488,0.9269835],"study_design_scores_gemma":[0.0001192556,0.0008585921,0.009645897,0.0003756686,0.0002238573,0.0008609215,0.0006461414,0.7731953,0.01396688,0.1607994,0.03908219,0.0002258878],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.05576851,0.002874844,0.9330671,0.001101494,0.0001475169,0.0002038783,0.001934257,0.002138482,0.002763904],"genre_scores_gemma":[0.3215572,0.001425669,0.6704041,0.0001962202,0.0003089076,0.0005206198,0.0036505,0.0001968374,0.001739893],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.005335013,"threshold_uncertainty_score":0.01122314,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1032342966842447,"score_gpt":0.2800124430266422,"score_spread":0.1767781463423975,"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."}}