{"id":"W3195710474","doi":"10.1145/3469096.3469866","title":"Efficient clustering of short text streams using online-offline clustering","year":2021,"lang":"en","type":"article","venue":"","topic":"Data Stream Mining Techniques","field":"Computer Science","cited_by":11,"is_retracted":false,"has_abstract":true,"ca_institutions":"Dalhousie University","funders":"","keywords":"Computer science; Cluster analysis; Data mining; Document clustering; Task (project management); Information retrieval; Similarity (geometry); Data stream clustering; Theoretical computer science; Artificial intelligence; Correlation clustering; CURE data clustering algorithm","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":[0.00110667,0.001387719,0.001427685,0.003010107,0.0009702857,0.001582284,0.002061154,0.00108846,0.001884631],"category_scores_gemma":[0.004507635,0.0004742341,0.0009890741,0.002578272,0.0005354628,0.002532802,0.001402148,0.001111787,0.001912593],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0009488993,"about_ca_system_score_gemma":0.001318561,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00364344,"about_ca_topic_score_gemma":0.004280224,"domain_scores_codex":[0.9986405,0.0002191569,0.0001274579,0.000445401,0.0004365648,0.0001309662],"domain_scores_gemma":[0.997588,0.0007752565,0.0002701544,0.0003634677,0.0008499032,0.0001531717],"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":[0.0009028331,0.000576351,0.005362868,0.0003205301,0.000164284,0.0003739377,0.0004935961,0.2212372,0.03785775,0.008228625,0.01271523,0.7117668],"study_design_scores_gemma":[0.00002693331,0.00005939208,0.0008666004,0.000008079892,0.00001382795,0.0000806363,0.00009555209,0.9825581,0.008427764,0.005656137,0.002187822,0.00001921931],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.01838819,0.0002025297,0.9780642,0.0001329899,0.0000782631,0.0001734413,0.0003453777,0.001904507,0.0007105629],"genre_scores_gemma":[0.2035352,0.0003002049,0.7869699,0.0001297113,0.0002339131,0.000399094,0.003287776,0.0004230609,0.004721082],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.00364344,"threshold_uncertainty_score":0.007244468,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.04463183452514439,"score_gpt":0.3083046983963606,"score_spread":0.2636728638712162,"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."}}