{"id":"W2162954778","doi":"10.1109/ssdbm.2007.32","title":"Effective Summarization of Multi-Dimensional Data Streams for Historical Stream Mining","year":2007,"lang":"en","type":"article","venue":"International Conference on Scientific and Statistical Database Management","topic":"Data Management and Algorithms","field":"Computer Science","cited_by":5,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Alberta","funders":"","keywords":"Automatic summarization; Computer science; Data stream; Data stream mining; Data mining; ENCODE; Data stream clustering; Cluster analysis; STREAMS; Task (project management); Space (punctuation); Simple (philosophy); Information retrieval; Artificial intelligence","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":[0.002103032,0.0007973162,0.001410952,0.002254276,0.000578806,0.001641571,0.001084412,0.0006463213,0.0006873194],"category_scores_gemma":[0.009483124,0.0004100978,0.0005057626,0.003022768,0.0003569902,0.003582865,0.0008448054,0.0006884742,0.0003655292],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0004837156,"about_ca_system_score_gemma":0.0006739002,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0008853615,"about_ca_topic_score_gemma":0.00154551,"domain_scores_codex":[0.9989858,0.000279849,0.000154159,0.0002040934,0.0003161859,0.00005992208],"domain_scores_gemma":[0.996492,0.001464935,0.0006209451,0.0005764302,0.0007015365,0.0001441807],"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":[0.0006403564,0.0003208691,0.009650347,0.0009541227,0.0002468042,0.0004001922,0.0008795037,0.2432955,0.03402634,0.0192099,0.008485946,0.6818901],"study_design_scores_gemma":[0.00003736197,0.00034304,0.002083944,0.00005991097,0.0001158663,0.0002315933,0.0003002361,0.958502,0.01497332,0.01635099,0.006968865,0.0000328424],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.0677231,0.002651316,0.9265907,0.0005571908,0.00009992992,0.0001582132,0.0007017258,0.0009714156,0.0005464468],"genre_scores_gemma":[0.4114531,0.002307803,0.5817508,0.0001036067,0.0003646742,0.0002170866,0.002569032,0.000136212,0.001097655],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.002254276,"threshold_uncertainty_score":0.01112205,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.09385673441405702,"score_gpt":0.3485527483789883,"score_spread":0.2546960139649312,"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."}}