{"id":"W2072732320","doi":"10.1007/s10618-005-0019-1","title":"Data Clustering with Partial Supervision","year":2006,"lang":"en","type":"article","venue":"Data Mining and Knowledge Discovery","topic":"Advanced Clustering Algorithms Research","field":"Computer Science","cited_by":75,"is_retracted":false,"has_abstract":false,"ca_institutions":"University of Alberta","funders":"","keywords":"Cluster analysis; Computer science; Artificial intelligence; Classifier (UML); Pattern recognition (psychology); Fuzzy clustering; Machine learning; Support vector machine; Data mining","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":"codex-gemma-dda1882f352a","candidate_categories":["open_science"],"consensus_categories":[],"category_scores_codex":[0.00049023,0.000178341,0.0001851581,0.00009239249,0.0002292482,0.0007158904,0.002913614,0.00004533082,0.000002409113],"category_scores_gemma":[0.00006487053,0.0001437177,0.000009930117,0.0003011455,0.0001064767,0.004600593,0.008966632,0.0001318614,0.00001554204],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00002082783,"about_ca_system_score_gemma":0.0001192684,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0001279364,"about_ca_topic_score_gemma":0.0005803453,"domain_scores_codex":[0.9980063,0.00006479952,0.0002096762,0.001044509,0.0002608342,0.0004138588],"domain_scores_gemma":[0.996532,0.0001714681,0.00004756732,0.003121869,0.00004051541,0.00008657637],"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":[0.0001302603,0.000413185,0.004966421,0.0002755834,0.00007413044,0.0002829958,0.001348295,0.0005931273,0.001118478,0.001369053,0.01762249,0.971806],"study_design_scores_gemma":[0.0006023248,0.0000972605,0.001134228,0.0001531405,0.00001020993,0.00008068766,0.0001850097,0.9664958,0.0001651092,0.00004569874,0.03072195,0.0003086045],"study_design_candidate":"design_other","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.04754636,0.001228566,0.9485965,0.0001599947,0.0002542332,0.0001182578,0.0005409045,0.0001489676,0.001406255],"genre_scores_gemma":[0.6104358,0.0001391495,0.3819658,0.00005243966,0.0008981257,0.00001681079,0.003573437,0.00006011574,0.002858275],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.9714974,"threshold_uncertainty_score":0.9990487,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.07186957610495115,"score_gpt":0.3322306622859221,"score_spread":0.260361086180971,"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."}}