{"id":"W4303986904","doi":"10.3897/bdj.10.e86089","title":"The disambiguation of people names in biological collections","year":2022,"lang":"en","type":"article","venue":"Biodiversity Data Journal","topic":"Genetic diversity and population structure","field":"Biochemistry, Genetics and Molecular Biology","cited_by":18,"is_retracted":false,"has_abstract":true,"ca_institutions":"Agriculture and Agri-Food Canada; McGill University","funders":"Horizon 2020 Framework Programme; Fonds Wetenschappelijk Onderzoek; Vlaamse regering; European Commission; Vlaamse Overheid; European Cooperation in Science and Technology; National Science Foundation","keywords":"Computer science; Information retrieval; Natural language processing; 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":[0.01548903,0.00082074,0.001193384,0.01836514,0.005686802,0.00694442,0.003380141,0.002239255,0.002225546],"category_scores_gemma":[0.05411947,0.001064953,0.001373191,0.01551679,0.004659411,0.01410424,0.01792616,0.002660018,0.002843076],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.002187383,"about_ca_system_score_gemma":0.005219227,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01172545,"about_ca_topic_score_gemma":0.01672752,"domain_scores_codex":[0.9787828,0.008360745,0.003000986,0.005146761,0.003790502,0.0009182215],"domain_scores_gemma":[0.9620125,0.01293525,0.005917569,0.0127199,0.005106564,0.001308321],"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":[0.0006857918,0.0002535152,0.1132877,0.00300027,0.0004902653,0.002552681,0.07187154,0.007076189,0.01480685,0.07959459,0.06214583,0.6442349],"study_design_scores_gemma":[0.00006841977,0.0001126753,0.05854923,0.0020892,0.0003821839,0.003160587,0.04018566,0.02224497,0.02836193,0.1036239,0.7407727,0.0004486012],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.3212095,0.01338702,0.5927456,0.006634805,0.003582205,0.001304665,0.01210948,0.008828792,0.04019794],"genre_scores_gemma":[0.346097,0.003550939,0.6192142,0.001422337,0.0005889401,0.0005510669,0.01908476,0.001618016,0.007872623],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.01836514,"threshold_uncertainty_score":0.08191478,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.04375251760692121,"score_gpt":0.2528607193802292,"score_spread":0.2091082017733081,"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."}}