{"id":"W3098382762","doi":"10.3897/biss.4.58482","title":"Improving Taxonomic Name Finding in the Biodiversity Heritage Library","year":2020,"lang":"en","type":"article","venue":"Biodiversity Information Science and Standards","topic":"Species Distribution and Climate Change","field":"Environmental Science","cited_by":2,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"","keywords":"World Wide Web; Index (typography); Biodiversity; Realm; Computer science; Quarter (Canadian coin); Taxon; Geography; Library science; Information retrieval; History; Biology; Archaeology; Ecology","routes":{"ca_aff":false,"ca_fund":false,"ca_venue":false,"about_ca":true,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.04528565,0.00116482,0.002418508,0.01885278,0.006144271,0.01900664,0.006648547,0.003551043,0.01106243],"category_scores_gemma":[0.1664833,0.001522724,0.001894272,0.01980804,0.003175785,0.0334226,0.01449867,0.003537195,0.01663347],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.006362675,"about_ca_system_score_gemma":0.01372039,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01773173,"about_ca_topic_score_gemma":0.01921403,"domain_scores_codex":[0.9532797,0.01684541,0.005215975,0.005165157,0.01755661,0.001937054],"domain_scores_gemma":[0.8570284,0.04718653,0.009705524,0.02993345,0.05194673,0.00419944],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"not_applicable","study_design_scores_codex":[0.0002826569,0.0004806222,0.02672865,0.0009490621,0.0001232594,0.0003414633,0.003843509,0.004857238,0.01394057,0.02760063,0.05917207,0.8616803],"study_design_scores_gemma":[0.0003117263,0.0005812501,0.02721043,0.002292542,0.0004106939,0.002434569,0.01439857,0.250721,0.1002652,0.09415602,0.5063422,0.000875761],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.05467648,0.002429369,0.8481145,0.01033357,0.0008895723,0.0008733331,0.002340941,0.0480484,0.03229384],"genre_scores_gemma":[0.05878035,0.0006605688,0.9287825,0.0009948611,0.0001283183,0.0001762509,0.003347898,0.002047837,0.005081441],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.04528565,"threshold_uncertainty_score":0.2394963,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02885904720582254,"score_gpt":0.2164791996118864,"score_spread":0.1876201524060638,"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."}}