{"id":"W2997432006","doi":"10.18280/ria.330608","title":"Design and Application of a Text Clustering Algorithm Based on Parallelized K-Means Clustering","year":2019,"lang":"en","type":"article","venue":"Revue d intelligence artificielle","topic":"Medical Research and Treatments","field":"Medicine","cited_by":21,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Inner Mongolia University of Technology; Inner Mongolia University","keywords":"Cluster analysis; Computer science; CURE data clustering algorithm; Correlation clustering; Canopy clustering algorithm; Single-linkage clustering; Algorithm; Data mining; Artificial intelligence","routes":{"ca_aff":false,"ca_fund":false,"ca_venue":true,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0005351887,0.0007469278,0.001078878,0.001622717,0.001510207,0.001292311,0.001700528,0.0009733123,0.002160705],"category_scores_gemma":[0.001525845,0.0004065204,0.000911543,0.002139549,0.0005939323,0.00145992,0.001039119,0.0008155322,0.001661909],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001097401,"about_ca_system_score_gemma":0.002564254,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.007248551,"about_ca_topic_score_gemma":0.005909335,"domain_scores_codex":[0.9989925,0.00008284749,0.0000682558,0.0004296426,0.0003361681,0.00009052219],"domain_scores_gemma":[0.9992262,0.00008074719,0.00005028182,0.0001002551,0.0004859514,0.00005654652],"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":[0.0004276729,0.0002429227,0.003100473,0.0002909247,0.000154508,0.0003246189,0.0004195968,0.1832594,0.07034364,0.01463266,0.02131939,0.7054842],"study_design_scores_gemma":[0.00005658397,0.00007350742,0.0008083257,0.00000839515,0.00002166944,0.000217083,0.0001198042,0.956369,0.02532543,0.005744762,0.01121776,0.00003766126],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.008533743,0.00008254057,0.9854218,0.0001408989,0.0001077156,0.0001908864,0.0001376011,0.003710147,0.001674748],"genre_scores_gemma":[0.08196037,0.0001062341,0.9140957,0.0001049186,0.00005927481,0.0003036991,0.0005820541,0.000224742,0.002563011],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.007248551,"threshold_uncertainty_score":0.0144127,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.04742661820200791,"score_gpt":0.32537590764925,"score_spread":0.2779492894472421,"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."}}