{"id":"W2047348798","doi":"10.1145/2213836.2213970","title":"Towards scalable summarization and visualization of large text corpora (abstract only)","year":2012,"lang":"en","type":"article","venue":"","topic":"Data Visualization and Analytics","field":"Computer Science","cited_by":1,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Alberta","funders":"","keywords":"Computer science; Automatic summarization; Scalability; Information retrieval; NoSQL; World Wide Web; Visualization; Analytics; Full text search; Data visualization; Data science; Search engine; Database; 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":[0.002183386,0.002700339,0.001365489,0.006547138,0.001190055,0.007215873,0.00190129,0.001226635,0.0136984],"category_scores_gemma":[0.01208963,0.001042434,0.001345127,0.005765178,0.000627287,0.005079406,0.004256347,0.002370219,0.01222763],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0007399889,"about_ca_system_score_gemma":0.001134417,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002863488,"about_ca_topic_score_gemma":0.003637974,"domain_scores_codex":[0.9982396,0.0004072881,0.0001968397,0.0004564156,0.0005978087,0.0001019214],"domain_scores_gemma":[0.9923511,0.002474579,0.0006785059,0.001416553,0.002676522,0.000402669],"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":[0.0007252409,0.000216654,0.001953394,0.002150856,0.0002789961,0.0006873354,0.001726258,0.01003999,0.05350539,0.008497402,0.3871745,0.533044],"study_design_scores_gemma":[0.0004036694,0.0003449972,0.007459065,0.0006464315,0.0002441726,0.0007818412,0.001694974,0.4459433,0.09668006,0.05716121,0.3883396,0.0003006234],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.01401573,0.002115366,0.6715355,0.002683676,0.000918137,0.000667999,0.04970301,0.2527768,0.005583787],"genre_scores_gemma":[0.06252534,0.001911174,0.798893,0.0004887283,0.0008784767,0.001354034,0.115732,0.01225845,0.005958763],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.0136984,"threshold_uncertainty_score":0.04582572,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02637914321600676,"score_gpt":0.3125271263731714,"score_spread":0.2861479831571646,"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."}}