{"meta":{"page":1,"per_page":50,"max_per_page":100,"total":1,"total_is_capped":false,"direct_labels_cover":0,"predictions_cover":1,"direct_label_status":"direct model label, unvalidated","prediction_status":"machine_predicted_unvalidated (Codex and Gemma teacher distillation)","score_status":"score_only:v0-immature-baseline (scores rank; they never assert a category)","snapshot":{"source":"OpenAlex, pinned release, all 482 partitions","release":"2026-06-24","frame_built":"2026-07-12","author_layer_release":"2026-06-26"},"query_hash":"1f1df8ae2f44","filters":{"venue":"LDV-Forum/Journal for language technology and computational linguistics"}},"results":[{"id":"W4381488486","doi":"10.21248/jlcl.36.2023.243","title":"Kencorpus: A Kenyan Language Corpus of Swahili, Dholuo and Luhya for Natural Language Processing Tasks","year":2023,"lang":"en","type":"article","venue":"LDV-Forum/Journal for language technology and computational linguistics","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":13,"is_retracted":false,"has_abstract":true,"routes":{"ca_aff":false,"ca_fund":true,"ca_venue":false,"about_ca":false},"ca_institutions":"","funders":"International Development Research Centre; Rockefeller Foundation","keywords":"Swahili; Computer science; Natural language processing; Artificial intelligence; Machine translation; Question answering; Linguistics","authors":[{"name":"Barack Wanjawa","is_ca":false},{"name":"Lilian Wanzare","is_ca":false},{"name":"Florence Indede","is_ca":false},{"name":"Owen McOnyango","is_ca":false},{"name":"Edward Ombui","is_ca":false},{"name":"Lawrence Muchemi","is_ca":false}],"retraction":null,"screen_n_in":null,"score":{"opus":0.008866185083867504,"gpt":0.2992442199912029,"spread":0.2903780349073354,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":["metaepi_narrow"],"consensus_categories":[],"category_scores_codex":[0.0005457808,0.0003261943,0.000466881,0.0009698561,0.0005412476,0.0002084247,0.0007951477,0.0003059944,0.000002085134],"category_scores_gemma":[0.001913699,0.0002930233,0.0001184763,0.0009747943,0.0003113989,0.0001805922,0.0003802124,0.0005780438,0.000001652511],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00004982826,"about_ca_system_score_gemma":0.0001637983,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00001513202,"about_ca_topic_score_gemma":0.00001691347,"domain_scores_codex":[0.9979501,0.00002841552,0.0005930857,0.0005075664,0.0003399978,0.0005808683],"domain_scores_gemma":[0.9979023,0.0004374803,0.0005087864,0.0002857074,0.0007454929,0.0001201749],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"theoretical_or_conceptual","study_design_scores_codex":[0.0002445506,0.0001764377,0.002482952,0.001115403,0.0002199875,0.0005415358,0.01157703,0.0003171479,0.007128544,0.2893246,0.002555193,0.6843166],"study_design_scores_gemma":[0.003811809,0.001256972,0.0003959933,0.0009410254,0.0001926504,0.002184733,0.01315246,0.3828782,0.01786248,0.5720155,0.00386883,0.001439292],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.221403,0.07609899,0.6938045,0.002758021,0.001627344,0.001375418,0.0003015249,0.002480704,0.0001505023],"genre_scores_gemma":[0.6835734,0.00003507625,0.315587,0.0002324467,0.00020074,0.0000381383,0.00007014895,0.00003216642,0.0002308906],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.6828773,"threshold_uncertainty_score":0.9999522,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null}]}