{"id":"W2071416419","doi":"10.1007/s11390-008-9115-1","title":"Clustering Text Data Streams","year":2008,"lang":"en","type":"article","venue":"Journal of Computer Science and Technology","topic":"Data Stream Mining Techniques","field":"Computer Science","cited_by":57,"is_retracted":false,"has_abstract":false,"ca_institutions":"","funders":"Simon Fraser University; Chinese University of Hong Kong; University of Hong Kong; Sun Yat-sen University","keywords":"Computer science; Cluster analysis; Data mining; Data stream mining; Data stream clustering; Smoothing; Document clustering; Fuzzy clustering; Semantics (computer science); Context (archaeology); Information retrieval; CURE data clustering algorithm; Artificial intelligence","routes":{"ca_aff":false,"ca_fund":true,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":["open_science"],"consensus_categories":[],"category_scores_codex":[0.00100642,0.0001143577,0.0002288988,0.0009693207,0.000244663,0.000151062,0.006268324,0.00007188895,0.000001153525],"category_scores_gemma":[0.0001384584,0.00009210185,0.00001685018,0.001611306,0.0009135813,0.002421502,0.004355715,0.0002591946,0.000003741926],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00003699298,"about_ca_system_score_gemma":0.000381945,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.000004105707,"about_ca_topic_score_gemma":0.000001596516,"domain_scores_codex":[0.9984441,0.00001612282,0.0003494718,0.0003922626,0.0005034505,0.0002946364],"domain_scores_gemma":[0.9978046,0.00004904453,0.0002853108,0.001332598,0.0004229986,0.0001054354],"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":[0.000002188936,0.00005157555,0.001953748,0.000004183909,0.0000101737,0.0002941315,0.0001561203,0.000005431815,0.00132763,0.005605311,0.005541744,0.9850478],"study_design_scores_gemma":[0.001394882,0.003043954,0.007718349,0.0003052738,0.00002460721,0.05636745,0.00007683987,0.8330725,0.02548843,0.02166189,0.04995553,0.0008902934],"study_design_candidate":"design_other","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.07902014,0.000182051,0.9171684,0.002893626,0.0003628439,0.00004623554,0.000001798171,0.0001783689,0.0001465336],"genre_scores_gemma":[0.4818583,0.0001135073,0.5178334,0.000125687,0.00006089588,3.999756e-7,2.502643e-7,0.000002962461,0.000004713998],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.9841574,"threshold_uncertainty_score":0.9991083,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03598818950116101,"score_gpt":0.2789750984262266,"score_spread":0.2429869089250656,"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."}}