{"id":"W2054808906","doi":"10.1145/2494266.2494279","title":"Interactive text document clustering using feature labeling","year":2013,"lang":"en","type":"article","venue":"","topic":"Text and Document Classification Technologies","field":"Computer Science","cited_by":15,"is_retracted":false,"has_abstract":true,"ca_institutions":"Dalhousie University","funders":"Natural Sciences and Engineering Research Council of Canada","keywords":"Computer science; Cluster analysis; Term (time); Discriminative model; Document clustering; Heuristic; Selection (genetic algorithm); Artificial intelligence; Cluster (spacecraft); Information retrieval; Data mining; Pattern recognition (psychology)","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":[0.00128084,0.001237514,0.001091844,0.004368111,0.001245937,0.001577276,0.002405572,0.001177389,0.004436443],"category_scores_gemma":[0.005074704,0.0004352842,0.001154208,0.004658341,0.0006622046,0.002518507,0.00158492,0.001191166,0.003336972],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0009473992,"about_ca_system_score_gemma":0.001130521,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.003041124,"about_ca_topic_score_gemma":0.005133945,"domain_scores_codex":[0.9980183,0.0003812737,0.0001469724,0.000577391,0.0007564991,0.0001195828],"domain_scores_gemma":[0.9963355,0.001333465,0.0003080229,0.0006992617,0.001173472,0.0001503349],"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":[0.0004890632,0.0003649002,0.002100468,0.0004354293,0.0001243542,0.0002289941,0.0006245404,0.01412402,0.08649924,0.006520201,0.02316175,0.8653271],"study_design_scores_gemma":[0.0002334396,0.000323163,0.003149117,0.00007884717,0.0001700247,0.00112162,0.0004306094,0.7407502,0.1496813,0.02961799,0.07422253,0.0002211108],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.00955676,0.0002397942,0.9796544,0.000122317,0.00005815651,0.000212994,0.000436503,0.008440433,0.001278714],"genre_scores_gemma":[0.05250287,0.0001205925,0.9412844,0.00009332077,0.00007919517,0.0003995995,0.001866549,0.0006882863,0.002965295],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.004436443,"threshold_uncertainty_score":0.01484138,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01998905261782122,"score_gpt":0.2722645220177291,"score_spread":0.2522754693999079,"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."}}