{"id":"W2096006132","doi":"10.1109/dest.2013.6611347","title":"Unstructured data extraction in distributed NoSQL","year":2013,"lang":"en","type":"article","venue":"","topic":"Semantic Web and Ontologies","field":"Computer Science","cited_by":1,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Saskatchewan","funders":"","keywords":"NoSQL; Computer science; Unstructured data; USable; Knowledge extraction; Big data; Data mining; Schema (genetic algorithms); Hidden Markov model; Information retrieval; Database; Artificial intelligence; 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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.003558985,0.0006258768,0.001106104,0.002499811,0.00115704,0.003885525,0.001737047,0.0007497433,0.002705469],"category_scores_gemma":[0.01062619,0.0007083683,0.001288043,0.003387648,0.0008222939,0.004748669,0.004736704,0.001199582,0.00139635],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0009666369,"about_ca_system_score_gemma":0.001722861,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.004300196,"about_ca_topic_score_gemma":0.004823224,"domain_scores_codex":[0.9958152,0.0006318275,0.0006999308,0.000783379,0.001848283,0.0002213809],"domain_scores_gemma":[0.9953181,0.001501939,0.0003063649,0.00183279,0.0008623897,0.0001785352],"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":[0.001137546,0.0005815332,0.01550944,0.001351253,0.0003616591,0.002514015,0.003637839,0.04275751,0.03128431,0.1016665,0.02995286,0.7692454],"study_design_scores_gemma":[0.000233685,0.0001563671,0.00496763,0.0002371677,0.000143747,0.001057438,0.001787585,0.6037813,0.05717202,0.2051395,0.1251599,0.0001636995],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.01913432,0.0002694569,0.9474311,0.00051918,0.00006855006,0.0003976021,0.002800735,0.02621727,0.003161769],"genre_scores_gemma":[0.2258035,0.0004720413,0.7522312,0.0004444825,0.00006903651,0.0004270819,0.01307074,0.00175123,0.005730756],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.004300196,"threshold_uncertainty_score":0.0188219,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03776107175899748,"score_gpt":0.2852752105419238,"score_spread":0.2475141387829264,"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."}}