{"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":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0002297805,0.0001849076,0.0002770821,0.0001370214,0.00006471638,0.0001062561,0.0008738894,0.00006972335,0.00003628226],"category_scores_gemma":[0.00007039431,0.0001794157,0.0000832281,0.0005229778,0.00004548033,0.0001683662,0.002052322,0.0001274236,0.000003664505],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00007025467,"about_ca_system_score_gemma":0.0001155997,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0001358853,"about_ca_topic_score_gemma":0.0001716608,"domain_scores_codex":[0.9983421,0.00004801068,0.0004487193,0.0005115515,0.0003294566,0.0003201753],"domain_scores_gemma":[0.998477,0.00007440106,0.00008637581,0.001112929,0.0001597933,0.00008949347],"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":[0.00002251885,0.00131648,0.002967985,0.0003094922,0.0001381185,0.0004479182,0.001325191,0.1923272,0.1618992,0.004012121,0.0007248962,0.6345089],"study_design_scores_gemma":[0.0001065629,0.00004273566,0.0003439657,0.0001394705,0.000008724656,0.0001016596,0.00009192338,0.952206,0.04648907,0.00003878304,0.0002423561,0.0001887138],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.1880812,0.00004670518,0.8095092,0.00009011327,0.0001833868,0.0000825152,0.0000211469,0.0003361642,0.001649595],"genre_scores_gemma":[0.4591964,0.000004925893,0.5405852,0.00006714678,0.00003540004,0.000001672671,0.00001411995,0.00001148387,0.00008362402],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.7598789,"threshold_uncertainty_score":0.7316359,"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."}}