{"id":"W4297231914","doi":"10.16995/dscn.8110","title":"Multilingual Research Projects: Non-Latin Script Challenges for Making Use of Standards, Authority Files, and Character Recognition","year":2022,"lang":"en","type":"article","venue":"Digital Studies / Le champ numérique","topic":"Digital Humanities and Scholarship","field":"Arts and Humanities","cited_by":6,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"","keywords":"Digitization; Computer science; Metadata; Newspaper; Natural language processing; Optical character recognition; Artificial intelligence; XML; Character (mathematics); Alphabet; Process (computing); Latin Americans; Information retrieval; Linguistics; World Wide Web; Telecommunications; Sociology; Media studies","routes":{"ca_aff":false,"ca_fund":false,"ca_venue":true,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":["sts"],"consensus_categories":[],"category_scores_codex":[0.001263772,0.0002406728,0.0004237443,0.0002072377,0.001398433,0.0007513915,0.0001690299,0.00004784405,0.00007451093],"category_scores_gemma":[0.0008621236,0.000229169,0.0001208252,0.00006177062,0.0005483053,0.001379141,0.0005117449,0.0004092048,0.00000187189],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001598625,"about_ca_system_score_gemma":0.0001162714,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00007982354,"about_ca_topic_score_gemma":0.0008514246,"domain_scores_codex":[0.997928,0.0001213612,0.0004419514,0.0004536266,0.0005894839,0.0004656095],"domain_scores_gemma":[0.9978004,0.0006297569,0.0001752347,0.0002005166,0.001138035,0.00005610065],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"qualitative","study_design_gemma":"qualitative","study_design_scores_codex":[0.0005281495,0.0004382263,0.000189675,0.0008379864,0.0002500799,0.00001973326,0.9229602,0.000002088909,0.0000226033,0.00774431,0.005012534,0.0619944],"study_design_scores_gemma":[0.0004667222,0.0006176308,0.0001169239,0.0001657667,6.155402e-7,0.000004160428,0.6648201,0.00002756118,0.00008402875,0.009073708,0.3243803,0.0002424948],"study_design_candidate":"qualitative","study_design_consensus":"qualitative","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9643928,0.002502967,0.000007159194,0.0004498768,0.0004680768,0.001109177,0.006543227,0.00007692727,0.02444977],"genre_scores_gemma":[0.9946508,0.0003510501,0.00001624479,0.00007248011,0.0004223325,0.000497731,0.0002135143,0.00004712262,0.003728703],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.3193678,"threshold_uncertainty_score":0.9999016,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.3801832065527054,"score_gpt":0.366869400317127,"score_spread":0.01331380623557837,"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."}}