{"id":"W4241938281","doi":"10.1145/376284.375672","title":"Data bubbles","year":2001,"lang":"en","type":"article","venue":"ACM SIGMOD Record","topic":"Advanced Clustering Algorithms Research","field":"Computer Science","cited_by":20,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of British Columbia","funders":"","keywords":"Cluster analysis; Computer science; Data mining; CURE data clustering algorithm; Data stream clustering; Canopy clustering algorithm; Hierarchical clustering; Correlation clustering; Set (abstract data type); Data set; Data compression; Key (lock); Constrained clustering; Determining the number of clusters in a data set; Fuzzy clustering; Algorithm; 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":["open_science"],"consensus_categories":[],"category_scores_codex":[0.0003168061,0.0001214837,0.0001353337,0.00009631546,0.0001231608,0.0001492538,0.006887774,0.00005068796,0.00005384663],"category_scores_gemma":[0.0005330507,0.0001119645,0.00002565907,0.0005068863,0.00004501903,0.001079359,0.005314135,0.0001992424,0.0002467739],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00004170333,"about_ca_system_score_gemma":0.00006051792,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00010036,"about_ca_topic_score_gemma":0.00006600292,"domain_scores_codex":[0.9983473,0.00005282237,0.0001757449,0.0006360096,0.0003520399,0.0004360795],"domain_scores_gemma":[0.9951127,0.0002581459,0.00004699509,0.004386411,0.00006503283,0.0001307079],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"not_applicable","study_design_scores_codex":[0.000007884198,0.00003175601,0.0007315229,0.000005157086,0.000009954611,0.00008758646,0.00005627658,0.00007121411,0.0004048741,0.0004059025,0.003807564,0.9943803],"study_design_scores_gemma":[0.0005280835,0.0001762575,0.003346164,0.00003320837,0.000003970431,0.0001944163,0.00002784676,0.4312353,0.0005222621,0.02460608,0.5388722,0.0004541536],"study_design_candidate":"design_other","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.01508502,0.0001595912,0.9793649,0.002498984,0.0005284749,0.0001286299,0.000008645049,0.0003343661,0.001891403],"genre_scores_gemma":[0.07812442,0.0004534129,0.9138129,0.0003588917,0.0004240146,0.00002405644,0.00002581275,0.00003255143,0.006743989],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.9939262,"threshold_uncertainty_score":0.9984854,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.123803096386827,"score_gpt":0.366511782884858,"score_spread":0.242708686498031,"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."}}