{"id":"W7154097441","doi":"10.5281/zenodo.19546907","title":"Format Identification in Context: Patterns & Hazards in Digital Preservation Workflows","year":2025,"lang":"","type":"article","venue":"Zenodo (CERN European Organization for Nuclear Research)","topic":"Digital and Traditional Archives Management","field":"Arts and Humanities","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Toronto Dementia Research Alliance; Library and Archives Canada","funders":"","keywords":"Digital preservation; Identification (biology); Workflow; Work (physics); Disparate system","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":["scholarly_communication"],"consensus_categories":[],"category_scores_codex":[0.02849351,0.0005014591,0.0005756793,0.005314136,0.005568343,0.01304519,0.002299692,0.001751416,0.002802851],"category_scores_gemma":[0.1005665,0.0007331843,0.0007777657,0.006794089,0.007227869,0.01521874,0.008511251,0.002077602,0.0008502153],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.003718298,"about_ca_system_score_gemma":0.008937951,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.008528087,"about_ca_topic_score_gemma":0.007144714,"domain_scores_codex":[0.9658661,0.01545482,0.004580835,0.004217355,0.008121556,0.001759358],"domain_scores_gemma":[0.898129,0.05266075,0.01776729,0.01973532,0.009671228,0.002036338],"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":[0.0005880817,0.0002705818,0.2042089,0.001414584,0.000100116,0.002333599,0.3093952,0.003929094,0.005725228,0.1029863,0.005527131,0.3635212],"study_design_scores_gemma":[0.00007553892,0.0003603403,0.1013649,0.003041676,0.000294265,0.00602653,0.3386336,0.02030638,0.01650191,0.3165674,0.1963733,0.0004541374],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.6476511,0.002724846,0.3068264,0.01024626,0.0002691018,0.0007334729,0.0008031568,0.001909475,0.02883618],"genre_scores_gemma":[0.8894634,0.0008723203,0.1044926,0.0004000445,0.00006617069,0.0002383207,0.000534975,0.0003843588,0.003547929],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.9869548,"threshold_uncertainty_score":0.1506899,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02847358283132882,"score_gpt":0.2230646387635029,"score_spread":0.194591055932174,"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."}}