{"id":"W2162148987","doi":"","title":"Clusterability: A Theoretical Study","year":2009,"lang":"en","type":"article","venue":"","topic":"Data Management and Algorithms","field":"Computer Science","cited_by":94,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Waterloo","funders":"","keywords":"Cluster analysis; Generality; Computer science; Pairwise comparison; Set (abstract data type); Property (philosophy); Data mining; Theoretical computer science; Artificial intelligence","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":[],"consensus_categories":[],"category_scores_codex":[0.0002887145,0.00005430112,0.00006150042,0.00003293614,0.00003944238,0.0002103299,0.0007069951,0.000008713888,0.00009171462],"category_scores_gemma":[0.00001369501,0.00003882143,0.00001954342,0.000170507,0.00002004255,0.0004108245,0.0002176414,0.00003930908,0.0001209558],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.000005714468,"about_ca_system_score_gemma":0.000004348112,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.000001718255,"about_ca_topic_score_gemma":8.532194e-7,"domain_scores_codex":[0.9993573,0.00003459194,0.00008857797,0.0002316717,0.0001588538,0.0001290443],"domain_scores_gemma":[0.9993793,0.00002036629,0.000009186921,0.0005388156,0.00001328127,0.00003906351],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"theoretical_or_conceptual","study_design_scores_codex":[8.877527e-7,0.0002590455,0.0002517033,3.235739e-7,0.000002617627,0.000008987085,0.0001275681,4.939085e-7,0.00000226502,0.8442063,0.001049453,0.1540904],"study_design_scores_gemma":[0.001379631,0.001914261,0.06152329,0.000004676641,0.00001495508,0.00000629249,0.0003849275,0.3069928,0.00009563561,0.6178994,0.009316547,0.0004676284],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":"theoretical_or_conceptual","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.01149149,0.00000353971,0.8928131,0.004831449,0.0001148775,0.0002346719,2.14753e-7,0.0002666713,0.09024399],"genre_scores_gemma":[0.9761204,4.136777e-7,0.02189187,0.001314742,0.00002731553,0.000002223728,5.300022e-7,9.853896e-7,0.0006415224],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.9646289,"threshold_uncertainty_score":0.2028216,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.00920970011625449,"score_gpt":0.2564770682758922,"score_spread":0.2472673681596377,"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."}}