{"id":"W4416701898","doi":"10.1016/j.envsoft.2025.106792","title":"Advancing geospatial data infrastructure in Dataverse via metadata automation, interactive tools and LLM case study","year":2025,"lang":"en","type":"article","venue":"Environmental Modelling & Software","topic":"Research Data Management Practices","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":false,"ca_institutions":"University of Toronto; Artificial Intelligence in Medicine (Canada)","funders":"Office of Advanced Cyberinfrastructure; Alfred P. Sloan Foundation; National Institutes of Health; National Science Foundation","keywords":"Geospatial analysis; Metadata; Geospatial metadata; Geospatial PDF; Metadata management; Meta Data Services; Reuse; Metadata modeling; Metadata repository","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":["metaresearch"],"consensus_categories":[],"category_scores_codex":[0.006663535,0.0003344835,0.0003013521,0.001752079,0.002087275,0.005183687,0.001965045,0.00230023,0.005095224],"category_scores_gemma":[0.009172161,0.0003407258,0.0005548301,0.003328261,0.001873603,0.006410554,0.005209832,0.001662304,0.001117239],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.002521056,"about_ca_system_score_gemma":0.002282594,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01245806,"about_ca_topic_score_gemma":0.01838009,"domain_scores_codex":[0.9966024,0.0015314,0.0001482266,0.0002181041,0.001052595,0.0004471422],"domain_scores_gemma":[0.99332,0.002895794,0.000212595,0.001970381,0.001034283,0.0005669265],"domain_codex":null,"domain_gemma":"reproducibility","domain_candidate":"reproducibility","domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"not_applicable","study_design_scores_codex":[0.001368661,0.002508993,0.03815496,0.001259943,0.00008002808,0.00888768,0.03455871,0.07083561,0.04375698,0.3550521,0.03066596,0.4128704],"study_design_scores_gemma":[0.0002442528,0.0008387399,0.01767274,0.0006991159,0.0001652628,0.002983459,0.03033356,0.2559438,0.1086027,0.07650866,0.5057796,0.0002281085],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"methods","genre_scores_codex":[0.6868526,0.0005463866,0.1666533,0.006330741,0.0001345287,0.0006700789,0.001374907,0.005271843,0.1321656],"genre_scores_gemma":[0.7871674,0.0003350131,0.1943533,0.0002260898,0.00003693934,0.0001518891,0.0009268636,0.0005054355,0.01629717],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.9933364,"threshold_uncertainty_score":0.03524053,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.04018958650219973,"score_gpt":0.3170950997399523,"score_spread":0.2769055132377526,"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."}}