{"id":"W6949619380","doi":"10.5281/zenodo.15241443","title":"Opening up translational data impact through the Data Citation Corpus","year":2025,"lang":"en","type":"article","venue":"Zenodo (CERN European Organization for Nuclear Research)","topic":"Biomedical Text Mining and Ontologies","field":"Biochemistry, Genetics and Molecular Biology","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"OpenAlex","funders":"National Center for Advancing Translational Sciences","keywords":"Metadata; Citation; Identifier; Identification (biology); Genomics; Translational research; Profiling (computer programming); Unique identifier; Data curation","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":[0.0006720811,0.00009110547,0.00007933295,0.00004294753,0.001076958,0.0003088433,0.002364512,0.00007444085,0.0007401306],"category_scores_gemma":[0.00113881,0.00006964756,0.00002442532,0.0002839908,0.0002195938,0.00002761239,0.002280551,0.0001472388,0.0002087153],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00001445228,"about_ca_system_score_gemma":0.00001545197,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00002765344,"about_ca_topic_score_gemma":0.00000163356,"domain_scores_codex":[0.9988165,0.0002170867,0.0001641067,0.0004263443,0.0001845707,0.0001914119],"domain_scores_gemma":[0.9985415,0.00003017843,0.00005559247,0.001166648,0.0001646289,0.00004150376],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"not_applicable","study_design_gemma":"not_applicable","study_design_scores_codex":[0.0001229736,0.00004752416,0.00002175418,0.00002121776,0.0001257287,0.000001802588,0.0003418537,0.00005419707,0.01261023,0.001205797,0.6544702,0.3309767],"study_design_scores_gemma":[0.0003696167,0.0000876719,0.001152698,0.000015742,0.00001818679,0.00002327126,0.000241721,0.001330475,0.0004262394,0.0002613709,0.9959899,0.00008309658],"study_design_candidate":"not_applicable","study_design_consensus":"not_applicable","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.093825,0.00716207,0.6559165,0.02679858,0.001504371,0.002000669,0.009372731,0.001119854,0.2023002],"genre_scores_gemma":[0.9484632,0.0002723662,0.002046689,0.0005049686,0.0002310625,5.284926e-8,0.04654154,0.0003054221,0.001634691],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.8546382,"threshold_uncertainty_score":0.8283199,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1238279532537489,"score_gpt":0.3570606096248361,"score_spread":0.2332326563710872,"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."}}