{"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":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0003721705,0.0001218984,0.0002618572,0.0001138729,0.00004371124,0.00001223965,0.00008777577,0.00006776235,0.0002799466],"category_scores_gemma":[0.0001344947,0.0001010299,0.00005544788,0.0001920549,0.00006950491,0.00003600576,0.00004080374,0.0001417794,0.0002967787],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00004555453,"about_ca_system_score_gemma":0.00005147322,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00003869947,"about_ca_topic_score_gemma":0.000002470016,"domain_scores_codex":[0.9988205,0.0000542378,0.0002963699,0.0003074606,0.0002708249,0.00025056],"domain_scores_gemma":[0.9989689,0.0003263918,0.00007090453,0.0003757307,0.00006828956,0.0001897649],"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":[0.0009840899,0.0006070625,0.001577626,0.0005958561,0.00006246952,0.00004106388,0.0004691308,0.109347,0.01862974,0.0001241602,0.00006348641,0.8674983],"study_design_scores_gemma":[0.0003951664,0.0007698327,0.0001652157,0.0002895798,0.00001809288,0.00001275612,0.0001145626,0.9763665,0.02113153,0.0000807605,0.0005768419,0.00007914421],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.01394018,0.0001134094,0.982763,0.0003680344,0.00004390812,0.0009981139,0.000002211308,0.00002813469,0.001743033],"genre_scores_gemma":[0.958486,0.00009814203,0.04015142,0.0001457804,0.00002832605,0.00006477591,0.00001300233,0.00001779429,0.0009947233],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.9445459,"threshold_uncertainty_score":0.4119878,"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."}}