{"id":"W2951446960","doi":"10.1145/3299869.3300094","title":"Fingerprints for Compressed Columnar Data Search","year":2019,"lang":"en","type":"article","venue":"","topic":"Data Management and Algorithms","field":"Computer Science","cited_by":2,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Waterloo","funders":"","keywords":"Computer science; Overhead (engineering); Block (permutation group theory); Data compression; Volume (thermodynamics); Compression (physics); Column (typography); Memory management; Compression ratio; Auxiliary memory; Data structure; Parallel computing; Minimax; Data compression ratio; Computer hardware; Semiconductor memory; Algorithm; Image compression; Artificial intelligence; Operating system; Computer network; Mathematics; Engineering","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":[0.0005265719,0.0004481676,0.0005908165,0.001582725,0.0004107099,0.001023851,0.000960238,0.0005227955,0.007368765],"category_scores_gemma":[0.004837676,0.0002699755,0.0002097351,0.00263767,0.0004822973,0.001861828,0.001173023,0.0004532294,0.001931105],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0005495965,"about_ca_system_score_gemma":0.001089953,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001241647,"about_ca_topic_score_gemma":0.001889304,"domain_scores_codex":[0.9993196,0.0001102018,0.00006229628,0.00009620465,0.0003455941,0.00006618629],"domain_scores_gemma":[0.9974295,0.0008989235,0.0002531988,0.0009204687,0.000397544,0.0001003614],"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":[0.0021419,0.0003118555,0.004113857,0.0005922319,0.00006204847,0.0003378158,0.0002415483,0.01745393,0.09623898,0.03022122,0.02307911,0.8252056],"study_design_scores_gemma":[0.0004142208,0.002191342,0.005419538,0.0002378659,0.0001047742,0.002948863,0.0004245155,0.6130416,0.2334872,0.05131571,0.09026826,0.0001461319],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.1066575,0.004665414,0.8627126,0.0006141113,0.0003107271,0.0005100327,0.003062686,0.01321305,0.008253957],"genre_scores_gemma":[0.4504392,0.001289429,0.5350754,0.0003849501,0.00015118,0.0003662836,0.00351643,0.000414762,0.008362442],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.007368765,"threshold_uncertainty_score":0.02465099,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.08198813717934422,"score_gpt":0.3090936588729068,"score_spread":0.2271055216935626,"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."}}