{"id":"W52853789","doi":"10.1137/1.9781611972764.40","title":"Collaborative Document Clustering","year":2006,"lang":"en","type":"article","venue":"","topic":"Peer-to-Peer Network Technologies","field":"Computer Science","cited_by":47,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Waterloo","funders":"","keywords":"Cluster analysis; Computer science; Document clustering; Data mining; Process (computing); Information retrieval; Cluster (spacecraft); Fuzzy clustering; Judgement; Similarity (geometry); Artificial intelligence; Computer network","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":[0.0001074176,0.00009139067,0.00009053432,0.00009157431,0.00006304959,0.0001824938,0.0007935631,0.0000374009,0.00001234638],"category_scores_gemma":[0.00001265525,0.00007852825,0.0000169052,0.0007724581,0.00002055354,0.0002631269,0.0006482998,0.0000542783,0.0001815501],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00006477195,"about_ca_system_score_gemma":0.00002461201,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0001003056,"about_ca_topic_score_gemma":0.0003017983,"domain_scores_codex":[0.9991669,0.00001377478,0.0001278781,0.0002635137,0.000182015,0.0002459059],"domain_scores_gemma":[0.9993986,0.00003052655,0.00002728599,0.0004450413,0.0000654278,0.00003305868],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"not_applicable","study_design_scores_codex":[0.000004907706,0.00004980314,0.0008219305,0.000005425058,0.0000155372,0.00005271599,0.0003045851,0.01316014,0.001772762,0.6758893,0.2182388,0.08968422],"study_design_scores_gemma":[0.001014043,0.0004839282,0.01922582,0.00006044678,0.000008226933,0.00004141869,0.0003315409,0.1053907,0.09143785,0.1688782,0.6116832,0.001444534],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.008527536,0.00005022351,0.9306142,0.007987333,0.0002298115,0.0001656231,5.332871e-7,0.001205514,0.05121925],"genre_scores_gemma":[0.5833477,0.000001943539,0.4124666,0.000443043,0.0000425588,0.00002652521,5.063217e-7,0.000004109159,0.003666975],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.5748202,"threshold_uncertainty_score":0.3202289,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.00511737081750828,"score_gpt":0.2290094822811698,"score_spread":0.2238921114636615,"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."}}