{"id":"W4395463230","doi":"10.18280/isi.290223","title":"Automatic Hindi OCR Error Correction Using MLM-BERT","year":2024,"lang":"fr","type":"article","venue":"Ingénierie des systèmes d information","topic":"Handwritten Text Recognition Techniques","field":"Computer Science","cited_by":2,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"","keywords":"Hindi; Computer science; Natural language processing; Artificial intelligence; Error detection and correction; Speech recognition; Algorithm","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":["metaepi_narrow","scholarly_communication","insufficient_payload"],"consensus_categories":["scholarly_communication"],"category_scores_codex":[0.001166574,0.0004452417,0.0003984073,0.000953974,0.0005414733,0.003433268,0.0005283771,0.0004238462,0.0006139953],"category_scores_gemma":[0.0004570203,0.0004860575,0.0002347604,0.001797159,0.0002978144,0.01805322,0.0002369399,0.0005215287,0.00219801],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001601215,"about_ca_system_score_gemma":0.0006389933,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0007523104,"about_ca_topic_score_gemma":0.00002921489,"domain_scores_codex":[0.9967536,0.0002509498,0.001293896,0.0003748701,0.0006274264,0.0006992097],"domain_scores_gemma":[0.9980889,0.0002600425,0.0004049779,0.0005070292,0.0005379047,0.0002011674],"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":[0.000005167451,0.00003680101,0.00005521537,0.001979342,0.00007934819,0.00003128357,0.01482246,0.0002179181,0.0001854637,0.007488244,0.01221375,0.962885],"study_design_scores_gemma":[0.0001671508,0.000170659,0.0004235805,0.004462497,0.00008483086,0.001356913,0.0005131741,0.934254,0.005175455,0.01512544,0.03770784,0.0005584824],"study_design_candidate":"design_other","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.04407404,0.00290545,0.9175894,0.0006378423,0.01608204,0.0008879452,0.00005046299,0.002969696,0.01480312],"genre_scores_gemma":[0.9133214,0.0002423016,0.08134677,0.0006464141,0.0006093681,0.0001499548,0.0001102784,0.00006330856,0.003510256],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.9623265,"threshold_uncertainty_score":0.9997591,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03190194894866177,"score_gpt":0.2796448416792766,"score_spread":0.2477428927306148,"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."}}