{"id":"W2990235080","doi":"10.33011/computel.v1i.345","title":"OCR Evaluation Tools for the 21st Century","year":2019,"lang":"en","type":"article","venue":"","topic":"Algorithms and Data Compression","field":"Computer Science","cited_by":17,"is_retracted":false,"has_abstract":true,"ca_institutions":"National Research Council Canada","funders":"","keywords":"Unicode; Computer science; Natural language processing; Word (group theory); Character (mathematics); Confusion; Artificial intelligence; Speech recognition; Linguistics; Psychology","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":[0.0004022403,0.00004486911,0.00004316805,0.00001313663,0.00008343648,0.0002239397,0.000531922,0.00001904714,0.0001585004],"category_scores_gemma":[0.00003646769,0.00002393898,0.00002977785,0.00007066075,0.00000557753,0.0006837207,0.0001728383,0.00003085014,0.0001569609],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00001362684,"about_ca_system_score_gemma":0.00003453259,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00001099293,"about_ca_topic_score_gemma":0.000001314182,"domain_scores_codex":[0.9994094,0.00001970501,0.00007439399,0.0001652353,0.0002268719,0.0001044173],"domain_scores_gemma":[0.9991211,0.0002542883,0.0000285273,0.0004973356,0.00007936262,0.00001940643],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0000016908,0.000009687134,0.00001777342,0.000002096283,0.000003408382,3.128647e-8,0.00005134768,0.00009952502,0.0001526974,0.1029739,0.004278559,0.8924093],"study_design_scores_gemma":[0.0002444807,0.00002693748,0.001188891,0.000004580673,0.00000344248,8.9077e-7,0.00007937913,0.7060381,0.0003358711,0.001560041,0.2904631,0.00005423152],"study_design_candidate":"design_other","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.001669571,0.0003798029,0.9776234,0.001317049,0.00131374,0.0008272907,0.000005286392,0.00007216896,0.01679171],"genre_scores_gemma":[0.8043029,0.0002424966,0.1891073,0.002769649,0.0004682779,0.000211201,0.00006969702,0.00001521279,0.002813176],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.892355,"threshold_uncertainty_score":0.2159456,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.04110869910776668,"score_gpt":0.2936906192732608,"score_spread":0.2525819201654942,"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."}}