{"id":"W2949063183","doi":"10.1109/icde.2019.00147","title":"Top-K Frequent Term Queries on Streaming Data","year":2019,"lang":"en","type":"article","venue":"","topic":"Data Management and Algorithms","field":"Computer Science","cited_by":11,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Alberta","funders":"","keywords":"Computer science; Term (time); Information retrieval; Streaming data; Social media; Data mining; World Wide Web","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":["insufficient_payload"],"consensus_categories":[],"category_scores_codex":[0.0001226928,0.0000748551,0.000068424,0.00004756252,0.00003447954,0.0002838713,0.001928345,0.00001357594,0.0001588259],"category_scores_gemma":[0.000006951989,0.00005773316,0.00001252906,0.0001014989,0.000009868243,0.001295004,0.001381369,0.00004432889,0.0008776459],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.000009989294,"about_ca_system_score_gemma":0.00001062863,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00004938276,"about_ca_topic_score_gemma":0.000006842919,"domain_scores_codex":[0.9992118,0.00001042484,0.00007541395,0.000375989,0.0001739691,0.0001523676],"domain_scores_gemma":[0.9981522,0.00002408698,0.00002562021,0.001759334,0.000008279098,0.00003046568],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"not_applicable","study_design_scores_codex":[0.000001589419,0.00007692047,0.004096609,0.00001646271,0.00001966908,0.00001620298,0.0001015927,0.000006683038,0.0001057172,0.4037092,0.0213652,0.5704841],"study_design_scores_gemma":[0.001205819,0.0004499863,0.04532708,0.0001272271,0.00001704016,0.000007017205,0.0004109369,0.2309418,0.004234581,0.004617844,0.7115198,0.001140953],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.1029499,0.00007359232,0.4934829,0.004378333,0.002696082,0.0005677271,0.00005553818,0.0008246662,0.3949712],"genre_scores_gemma":[0.7663463,0.00006092269,0.1708588,0.002636641,0.0002671936,0.000008886303,0.0003739432,0.00001979833,0.05942752],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.6901546,"threshold_uncertainty_score":0.9999003,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03112682957215243,"score_gpt":0.2643527266923595,"score_spread":0.2332258971202071,"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."}}