{"id":"W2021502500","doi":"10.1109/sysose.2013.6575236","title":"Terms extraction from unstructured data silos","year":2013,"lang":"en","type":"article","venue":"","topic":"Semantic Web and Ontologies","field":"Computer Science","cited_by":14,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Saskatchewan","funders":"","keywords":"NoSQL; Computer science; Unstructured data; Knowledge extraction; Big data; Semi-structured data; Schema (genetic algorithms); Data mining; Data modeling; Graph database; Graph; Database; Data science; Information retrieval; Relational database; Theoretical computer science","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":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.00004099856,0.00006594513,0.00007484281,0.00002410501,0.0000419922,0.0002077865,0.001317788,0.00004173312,0.0003857797],"category_scores_gemma":[0.0000347893,0.00004554524,0.00001318258,0.00006809981,0.00002013578,0.001269456,0.0003863036,0.00005452731,0.0004996857],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.000006399525,"about_ca_system_score_gemma":0.00001224206,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001636323,"about_ca_topic_score_gemma":0.00008896641,"domain_scores_codex":[0.9993377,0.00001666703,0.0001008744,0.0003027775,0.0001163644,0.0001256056],"domain_scores_gemma":[0.9986705,0.00007353276,0.00003348377,0.001167758,0.00001829521,0.00003646301],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"observational","study_design_scores_codex":[0.00000223173,0.00004716084,0.008383977,0.000004623422,0.00003758508,0.00001560324,0.0002885021,0.000006885457,0.02868558,0.02799921,0.08867906,0.8458496],"study_design_scores_gemma":[0.0004288245,0.00003788718,0.5711055,0.00001087946,0.00001083157,0.00003043977,0.0001671439,0.2288081,0.02001466,0.1643227,0.01467619,0.0003867673],"study_design_candidate":"design_other","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.1156515,0.00006930673,0.8574604,0.003652386,0.001151085,0.0001629172,0.000004848497,0.0004932521,0.02135432],"genre_scores_gemma":[0.8556886,0.000004701964,0.1434597,0.0004075148,0.00007898446,0.000002942513,0.000014628,0.000002216832,0.0003407347],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.8454628,"threshold_uncertainty_score":0.6422615,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.04294107636015791,"score_gpt":0.2775724917454833,"score_spread":0.2346314153853254,"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."}}