{"id":"W1974097983","doi":"10.1109/sose.2014.25","title":"Towards Knowledge Discovery in Big Data","year":2014,"lang":"en","type":"article","venue":"","topic":"Data Quality and Management","field":"Decision Sciences","cited_by":56,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Saskatchewan","funders":"","keywords":"NoSQL; Computer science; Unstructured data; Big data; Knowledge extraction; Data warehouse; Data science; Semi-structured data; Schema (genetic algorithms); SQL; Analytics; Database; Data mining; Information retrieval; Relational database","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":[0.0147842,0.0009675763,0.001883081,0.010824,0.001766814,0.01287265,0.003577845,0.002959574,0.002102733],"category_scores_gemma":[0.04476554,0.001252104,0.00246381,0.009714837,0.004471081,0.01562193,0.01145128,0.005248034,0.001688707],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001889918,"about_ca_system_score_gemma":0.004650117,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00257408,"about_ca_topic_score_gemma":0.003150753,"domain_scores_codex":[0.9887967,0.004394358,0.00092893,0.001142554,0.004388572,0.0003488983],"domain_scores_gemma":[0.9675885,0.02121168,0.001257638,0.00639064,0.002670078,0.0008814379],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"theoretical_or_conceptual","study_design_scores_codex":[0.0001499986,0.0002972066,0.005279253,0.002290882,0.000424802,0.001042302,0.003099963,0.02292797,0.00319296,0.5786696,0.01973736,0.3628877],"study_design_scores_gemma":[0.00002521243,0.0000265376,0.0005439337,0.0004595502,0.00005404349,0.0003029252,0.001104939,0.1065265,0.002683279,0.8372506,0.05098389,0.00003860679],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":"theoretical_or_conceptual","genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.007132181,0.004861862,0.9679893,0.01181101,0.00024049,0.000235239,0.0006991567,0.001570097,0.005460706],"genre_scores_gemma":[0.05874638,0.004662092,0.9308152,0.001633394,0.0002319026,0.0002910286,0.001767208,0.0001253348,0.001727466],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.0147842,"threshold_uncertainty_score":0.07818729,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.5125727714900005,"score_gpt":0.4766640422878515,"score_spread":0.03590872920214894,"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."}}