{"id":"W2130734227","doi":"10.1109/ccece.2004.1344986","title":"Evaluation of three dimensionality reduction techniques for document classification","year":2004,"lang":"en","type":"article","venue":"","topic":"Text and Document Classification Technologies","field":"Computer Science","cited_by":2,"is_retracted":false,"has_abstract":true,"ca_institutions":"Dalhousie University","funders":"","keywords":"Dimensionality reduction; Computer science; Curse of dimensionality; Reduction (mathematics); Artificial intelligence; Latent semantic analysis; Categorization; Document classification; Search engine indexing; Machine learning; Data mining; Representation (politics); Measure (data warehouse); Pattern recognition (psychology); Information retrieval; Mathematics","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":[0.009171806,0.001518761,0.001582715,0.004267805,0.0008014128,0.001592565,0.001234202,0.001172866,0.001175274],"category_scores_gemma":[0.01936793,0.0002619534,0.001643956,0.004426398,0.0006157397,0.002241021,0.001115358,0.001082101,0.0008124893],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001608747,"about_ca_system_score_gemma":0.001260586,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.004451024,"about_ca_topic_score_gemma":0.003130684,"domain_scores_codex":[0.9901919,0.002824158,0.0008475025,0.0008733671,0.004879836,0.0003831124],"domain_scores_gemma":[0.9837684,0.009154175,0.0006158308,0.001451195,0.004681189,0.0003291964],"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":[0.001060385,0.0008042589,0.004960262,0.0005321591,0.0004603736,0.00004277995,0.0002253903,0.02577037,0.007948123,0.00259107,0.005325255,0.9502795],"study_design_scores_gemma":[0.0004758323,0.003475967,0.01951409,0.000089673,0.0004387232,0.00045365,0.0004319628,0.9190415,0.04305643,0.00404952,0.00882796,0.0001447392],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.4726829,0.01004288,0.4865342,0.001434193,0.0005333087,0.002016734,0.002024883,0.009526306,0.01520468],"genre_scores_gemma":[0.3932874,0.002693938,0.5953243,0.0001250295,0.0001625788,0.0008621488,0.003806011,0.0002599681,0.00347863],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.009171806,"threshold_uncertainty_score":0.04850566,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.09872650376107284,"score_gpt":0.3484091910519327,"score_spread":0.2496826872908599,"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."}}