{"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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.001522176,0.001063996,0.001283575,0.009087631,0.001132074,0.002856066,0.001189881,0.001142245,0.002501728],"category_scores_gemma":[0.007556651,0.000525563,0.001432164,0.006772095,0.0004699196,0.002391671,0.001080394,0.001242978,0.002017163],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.000855826,"about_ca_system_score_gemma":0.001280292,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002108877,"about_ca_topic_score_gemma":0.002417153,"domain_scores_codex":[0.9980692,0.00022182,0.0002617652,0.0004511868,0.0008579333,0.0001381077],"domain_scores_gemma":[0.9967043,0.0009564135,0.0002618553,0.0004531812,0.001468165,0.0001560493],"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":[0.001097413,0.0006613533,0.01789875,0.0003734989,0.0003211978,0.0003657052,0.0003424887,0.03236965,0.02966657,0.008125515,0.01740556,0.8913723],"study_design_scores_gemma":[0.00009901564,0.0003092639,0.008876882,0.00007474307,0.000243672,0.0005521827,0.0004026192,0.9134007,0.0361388,0.02293777,0.0169031,0.00006127149],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.1405475,0.001880565,0.8430446,0.001184435,0.0009966471,0.001082576,0.003294742,0.004141348,0.003827539],"genre_scores_gemma":[0.4376804,0.002010755,0.5300859,0.0002645063,0.0009687713,0.0008120037,0.01408909,0.0003969062,0.0136916],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.009087631,"threshold_uncertainty_score":0.008369088,"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."}}