{"id":"W1881830221","doi":"10.1111/coin.12064","title":"Document Clustering With Dual Supervision Through Feature Reweighting","year":2015,"lang":"en","type":"article","venue":"Computational Intelligence","topic":"Text and Document Classification Technologies","field":"Computer Science","cited_by":22,"is_retracted":false,"has_abstract":true,"ca_institutions":"Dalhousie University","funders":"Canadian Network for Research and Innovation in Machining Technology, Natural Sciences and Engineering Research Council of Canada","keywords":"Cluster analysis; Computer science; Feature (linguistics); Brown clustering; Artificial intelligence; Pairwise comparison; Correlation clustering; Pattern recognition (psychology); Fuzzy clustering; Consensus clustering; Data mining; Conceptual clustering; Machine learning; Canopy clustering algorithm","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":[0.001972641,0.001129378,0.001354108,0.001568511,0.0008718021,0.001300877,0.002223541,0.001140459,0.002048929],"category_scores_gemma":[0.007341409,0.0004503411,0.0009166433,0.001985703,0.0008963191,0.002230774,0.001795163,0.001733244,0.00203392],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0007022878,"about_ca_system_score_gemma":0.001399643,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.003338062,"about_ca_topic_score_gemma":0.006517736,"domain_scores_codex":[0.9980367,0.0005080294,0.0001293478,0.000640941,0.0005578669,0.0001271553],"domain_scores_gemma":[0.9966827,0.0007871346,0.0002784317,0.001069853,0.00104232,0.0001396136],"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":[0.0004675131,0.0003269544,0.002924056,0.0002392727,0.0002018913,0.0001146513,0.0003843744,0.08059643,0.02527431,0.007424016,0.01203991,0.8700067],"study_design_scores_gemma":[0.00003825895,0.00008238704,0.0008642941,0.0000180656,0.00004503735,0.0001042751,0.00005572476,0.973424,0.01067291,0.01097782,0.003686925,0.00003037253],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.02320809,0.0002368133,0.9713222,0.0001413759,0.00004822919,0.000116448,0.0002059555,0.003385231,0.001335498],"genre_scores_gemma":[0.3400945,0.000192639,0.6529134,0.0001834969,0.0001182388,0.0003009667,0.001748711,0.0004732422,0.003974804],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.003338062,"threshold_uncertainty_score":0.01043248,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.05995408911021713,"score_gpt":0.3051056388127768,"score_spread":0.2451515497025597,"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."}}