{"id":"W3172225000","doi":"10.48550/arxiv.2106.09325","title":"Central Kurdish machine translation: First large scale parallel corpus\\n and experiments","year":2021,"lang":"en","type":"preprint","venue":"arXiv (Cornell University)","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":2,"is_retracted":false,"has_abstract":true,"ca_institutions":"Toronto Metropolitan University","funders":"","keywords":"Machine translation; Translation (biology); Computer science; Parallel corpora; Natural language processing; Scale (ratio); Artificial intelligence; Geography; Cartography; Chemistry","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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.004191843,0.001788145,0.001629229,0.002639718,0.003638596,0.002409246,0.002049905,0.001839798,0.01430833],"category_scores_gemma":[0.0139156,0.0007701497,0.0009877905,0.004445743,0.002363411,0.004554376,0.005035629,0.00271601,0.01307574],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001531257,"about_ca_system_score_gemma":0.002798986,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01245422,"about_ca_topic_score_gemma":0.02336294,"domain_scores_codex":[0.9953674,0.001877859,0.0004848225,0.001004103,0.0009115828,0.000354223],"domain_scores_gemma":[0.9906569,0.003413358,0.0002559785,0.002925724,0.002241609,0.0005064324],"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":[0.00306592,0.00431021,0.007574018,0.004805489,0.0006093754,0.003088568,0.003807638,0.01356741,0.03517069,0.008650361,0.4527588,0.4625915],"study_design_scores_gemma":[0.005007495,0.003696458,0.06157495,0.001092097,0.0008518674,0.006844803,0.01004809,0.1303125,0.1251833,0.02236152,0.6322759,0.0007509606],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.7175395,0.01283981,0.05047813,0.004931282,0.003158847,0.003796176,0.08468634,0.03929823,0.08327167],"genre_scores_gemma":[0.4959204,0.002245608,0.1242083,0.001583109,0.0005798168,0.002727393,0.348805,0.003925527,0.02000473],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.01430833,"threshold_uncertainty_score":0.04786617,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.04133679020691154,"score_gpt":0.2067070709236566,"score_spread":0.1653702807167451,"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."}}