{"id":"W1970107047","doi":"10.1109/icmlc.2010.5580499","title":"Incremental clustering algorithm based on phrase-semantic similarity histogram","year":2010,"lang":"en","type":"article","venue":"","topic":"Web Data Mining and Analysis","field":"Computer Science","cited_by":11,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Waterloo","funders":"","keywords":"Cluster analysis; Computer science; Fuzzy clustering; Correlation clustering; CURE data clustering algorithm; Canopy clustering algorithm; Document clustering; Data stream clustering; Data mining; Artificial intelligence; Clustering high-dimensional data; Brown clustering; Single-linkage clustering; Pattern recognition (psychology); Information retrieval","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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0006024535,0.000601951,0.001261802,0.003714728,0.00119928,0.00104186,0.002642533,0.0008642965,0.002786026],"category_scores_gemma":[0.002425805,0.0003330225,0.0009871912,0.003912885,0.0005508864,0.001933709,0.001044886,0.000856926,0.001388843],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0009030461,"about_ca_system_score_gemma":0.001680095,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.007879733,"about_ca_topic_score_gemma":0.006751334,"domain_scores_codex":[0.9991171,0.00008885498,0.00005709008,0.0002308716,0.0004278712,0.00007831202],"domain_scores_gemma":[0.9990098,0.0002063112,0.00007420717,0.0001194958,0.000538514,0.00005156693],"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":[0.0002362337,0.0002007995,0.002190175,0.0001770382,0.0001642613,0.0001728773,0.0002806319,0.06422837,0.01478869,0.01457751,0.01186167,0.8911217],"study_design_scores_gemma":[0.0000859577,0.0001515815,0.002279615,0.00001384548,0.0001022249,0.0004751625,0.0001464443,0.9552265,0.0124078,0.02005196,0.008967233,0.00009168335],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.01647888,0.0002873102,0.9791307,0.0001073436,0.0001064942,0.0002007835,0.0002146439,0.001830818,0.001643002],"genre_scores_gemma":[0.1665679,0.0003876197,0.8267474,0.0001196695,0.0001481709,0.0003618141,0.001556289,0.0002505163,0.003860625],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.007879733,"threshold_uncertainty_score":0.01566774,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01338385580906149,"score_gpt":0.243526709109979,"score_spread":0.2301428533009175,"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."}}