{"id":"W4407998460","doi":"10.1017/eds.2024.53","title":"OpenForest: a data catalog for machine learning in forest monitoring","year":2025,"lang":"en","type":"article","venue":"Environmental Data Science","topic":"Remote Sensing and LiDAR Applications","field":"Environmental Science","cited_by":12,"is_retracted":false,"has_abstract":true,"ca_institutions":"Université de Montréal; McGill University; Mila - Quebec Artificial Intelligence Institute","funders":"Institut de Valorisation des Données; Deutsche Forschungsgemeinschaft; Canadian Institute for Advanced Research","keywords":"Computer science; Data science; Forestry; Artificial intelligence; Machine learning; Geography","routes":{"ca_aff":true,"ca_fund":true,"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":["open_science"],"consensus_categories":[],"category_scores_codex":[0.005591738,0.001927758,0.001600745,0.01562315,0.001167953,0.003592812,0.003802828,0.002234239,0.02296119],"category_scores_gemma":[0.02512827,0.0009381889,0.001355501,0.01899506,0.0006548763,0.005921159,0.005688865,0.002363024,0.02692118],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001332665,"about_ca_system_score_gemma":0.003391909,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.007854529,"about_ca_topic_score_gemma":0.01307485,"domain_scores_codex":[0.9954773,0.0007427784,0.0009674178,0.000824757,0.001672043,0.0003157368],"domain_scores_gemma":[0.981483,0.00610737,0.00163775,0.006259335,0.00322662,0.00128573],"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":[0.0002068367,0.0001516228,0.006063292,0.002302613,0.0001137807,0.0001610548,0.0001623712,0.003351169,0.001350944,0.007429337,0.8960223,0.08268475],"study_design_scores_gemma":[0.0001429126,0.00003893978,0.008613291,0.0008135504,0.00003678221,0.0001779838,0.0001031485,0.007607458,0.001857156,0.01214746,0.9683737,0.00008761634],"study_design_candidate":"not_applicable","study_design_consensus":"not_applicable","genre_codex":"dataset","genre_gemma":"software","genre_scores_codex":[0.002351357,0.001451853,0.01489679,0.0004396679,0.0001614272,0.0003090581,0.949495,0.02590537,0.004989496],"genre_scores_gemma":[0.002919342,0.0004326124,0.01865119,0.0001643282,0.00004491904,0.0004673999,0.9747134,0.001928301,0.0006785382],"genre_candidate":"software","genre_consensus":null,"teacher_disagreement_score":0.9961972,"threshold_uncertainty_score":0.07681286,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03437498071578052,"score_gpt":0.3017214673339243,"score_spread":0.2673464866181438,"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."}}