{"id":"W4413775151","doi":"10.5121/csit.2025.151603","title":"Towards stable AI systems for Evaluating Arabic Pronunciations","year":2025,"lang":"en","type":"article","venue":"","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"York University; New York University Abu Dhabi","keywords":"Arabic; Computer science; Natural language processing; Artificial intelligence; Writing system; Linguistics; Philosophy","routes":{"ca_aff":false,"ca_fund":true,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.003623138,0.001746984,0.000773099,0.00229291,0.000617023,0.003219326,0.001911879,0.001425797,0.005098764],"category_scores_gemma":[0.0179816,0.0004478253,0.0005444302,0.001338139,0.0007855821,0.002700965,0.002133943,0.002191808,0.006078239],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001140222,"about_ca_system_score_gemma":0.00118134,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.006359845,"about_ca_topic_score_gemma":0.006936092,"domain_scores_codex":[0.9968214,0.001048928,0.0002410575,0.001022726,0.0006732433,0.0001925692],"domain_scores_gemma":[0.9946286,0.00198456,0.0003510869,0.0006609585,0.002184584,0.0001901189],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"bench_or_experimental","study_design_scores_codex":[0.0005806431,0.0002644836,0.008762614,0.0004214,0.0002679251,0.0001645035,0.0004681747,0.08994993,0.05052672,0.009268273,0.02077718,0.8185481],"study_design_scores_gemma":[0.00001854436,0.0001366832,0.002775536,0.00005332442,0.00003631921,0.0000734762,0.000222374,0.9593145,0.02329073,0.008734111,0.005303744,0.00004065542],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.05305629,0.001229349,0.9107507,0.0006988743,0.0005732985,0.000273672,0.002515285,0.02306997,0.007832663],"genre_scores_gemma":[0.4632691,0.0005970848,0.5170008,0.0004559588,0.0002065585,0.0006163644,0.008421835,0.0009371262,0.008495112],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.006359845,"threshold_uncertainty_score":0.01916122,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03002963023902807,"score_gpt":0.360582316062258,"score_spread":0.3305526858232299,"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."}}