{"id":"W4297878801","doi":"10.1590/scielopreprints.4755","title":"ARCHIVING AND LANGUAGE DOCUMENTATION","year":2022,"lang":"en","type":"preprint","venue":"","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":7,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Alberta","funders":"","keywords":"Documentation; Relevance (law); Focus (optics); Indigenous; Indigenous language; Library science; Linguistics; Political science; World Wide Web; Computer science","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":[0.0002151958,0.0001217531,0.0001131991,0.0001320806,0.00008475517,0.0004286017,0.0008389088,0.00004778019,0.0001164099],"category_scores_gemma":[0.00002830977,0.0001110727,0.00002969429,0.00008931885,0.00001751135,0.0002488978,0.004569974,0.0004559284,0.000002191519],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00004682208,"about_ca_system_score_gemma":0.00004295354,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0002229715,"about_ca_topic_score_gemma":0.000008551858,"domain_scores_codex":[0.9990796,0.00005816167,0.0001233372,0.0004067235,0.0002076947,0.0001244688],"domain_scores_gemma":[0.9993762,0.00005135579,0.00008904897,0.0004345643,0.00001378848,0.00003507528],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"theoretical_or_conceptual","study_design_scores_codex":[0.000004305143,0.00004096514,0.0004728831,0.0004211351,0.00003719713,0.0001404601,0.01395243,0.00003148771,0.005563847,0.343709,0.002496384,0.6331299],"study_design_scores_gemma":[0.0002667239,0.00009034754,0.0008388839,0.0002077781,0.00002447236,0.00008584173,0.0005750831,0.03229415,0.02406589,0.9385206,0.00186472,0.001165519],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.008801331,0.004867795,0.9777222,0.0009595062,0.0002724105,0.0002371095,0.000003785107,0.001431098,0.005704715],"genre_scores_gemma":[0.207486,0.00004016568,0.7909704,0.0003705269,0.00003259381,0.00004739166,0.00001667237,0.000008389093,0.001027863],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.6319644,"threshold_uncertainty_score":0.5696146,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01005173041471197,"score_gpt":0.3036480371038187,"score_spread":0.2935963066891067,"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."}}