{"meta":{"query_hash":"030217acec42","filters":{"venue":"Northern European Journal of Language Technology"},"cohort_total":2,"direct_labels_cover":0,"predictions_cover":2,"exported":2,"export_cap":100000,"truncated":false,"label_status":"direct model label, unvalidated","prediction_status":"machine_predicted_unvalidated (Codex and Gemma teacher distillation)","score_status":"score_only:v0-immature-baseline","snapshot":{"source":"OpenAlex, pinned release, all 482 partitions","release":"2026-06-24","frame_built":"2026-07-12"},"permalink":"https://metacan.xera.ac/q/030217acec42","api":"https://metacan.xera.ac/api/v1/cohort?venue=Northern+European+Journal+of+Language+Technology"},"results":[{"id":"W4311428537","doi":"10.3384/nejlt.2000-1533.2022.4315","title":"Part-of-Speech and Morphological Tagging of Algerian Judeo-Arabic","year":2022,"lang":"en","type":"article","venue":"Northern European Journal of Language Technology","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"route_ca_aff":false,"route_ca_fund":true,"route_ca_venue":false,"route_about_ca":false,"ca_institutions":"","funders":"Azrieli Foundation; Israel Science Foundation","keywords":"Computer science; Natural language processing; Annotation; Arabic; Preprocessor; Pipeline (software); Artificial intelligence; Part of speech; Linguistics","score_opus":0.010133932396906404,"score_gpt":0.23726246352599364,"score_spread":0.22712853112908724,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W4311428537","genre_codex":"empirical","genre_gemma":"empirical","domain_codex":null,"domain_gemma":null,"model_version":"metacan-v3-hybrid-931329e0061c","genre_candidate":"empirical","genre_consensus":"empirical","domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.951947,0.0008960935,0.026737956,0.00038590963,0.00021868387,0.00012472286,0.0061394116,0.0014614221,0.012088937],"genre_scores_gemma":[0.90054953,0.00048680865,0.07479942,0.0001019851,0.00010279026,0.00014423576,0.014365961,0.00035931516,0.00909005],"study_design_codex":"design_other","study_design_gemma":"bench_or_experimental","domain_scores_codex":[0.9994821,0.00015174753,0.000062828396,0.00016664427,0.00010098641,0.000035757697],"domain_scores_gemma":[0.9972621,0.0012015442,0.00029127634,0.0004148098,0.0007452202,0.00008502127],"candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0006144456,0.00038897927,0.00020173068,0.0026105125,0.0011415058,0.0007654896,0.0004077101,0.00049357984,0.0033584964],"category_scores_gemma":[0.0025351332,0.00019230414,0.000189204,0.0017045666,0.00052318914,0.00082727947,0.0005433368,0.00035952474,0.0022313676],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_system_candidate":false,"about_ca_system_consensus":false,"study_design_scores_codex":[0.0011929076,0.00032035023,0.05989053,0.001373649,0.00009209669,0.0037612293,0.014223553,0.008956791,0.29579237,0.0062351357,0.01867493,0.5894864],"study_design_scores_gemma":[0.00008433074,0.00026773722,0.39323506,0.0002942482,0.00014858009,0.0041103596,0.011457548,0.11674066,0.28890082,0.005413044,0.17913564,0.00021198511],"about_ca_topic_score_codex":0.004796105,"about_ca_topic_score_gemma":0.010171592,"teacher_disagreement_score":0.004796105,"about_ca_system_score_codex":0.00045128312,"about_ca_system_score_gemma":0.00050798553,"threshold_uncertainty_score":0.011235237},"labels":[],"label_agreement":null},{"id":"W4402487814","doi":"10.3384/nejlt.2000-1533.2024.5217","title":"Documenting Geographically and Contextually Diverse Language Data Sources","year":2024,"lang":"en","type":"article","venue":"Northern European Journal of Language Technology","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"route_ca_aff":false,"route_ca_fund":true,"route_ca_venue":false,"route_about_ca":false,"ca_institutions":"","funders":"Sorbonne Université; Università Bocconi; Centre National de la Recherche Scientifique; Grand Équipement National De Calcul Intensif; Simon Fraser University","keywords":"Spatial contextual awareness; Context (archaeology); Geography; Remote sensing","score_opus":0.011623514470331659,"score_gpt":0.2669204921008031,"score_spread":0.25529697763047143,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W4402487814","genre_codex":"methods","genre_gemma":"methods","domain_codex":null,"domain_gemma":null,"model_version":"metacan-v3-hybrid-931329e0061c","genre_candidate":"methods","genre_consensus":"methods","domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.14721452,0.0010864297,0.77889717,0.003677502,0.00016250415,0.0032951368,0.0376208,0.0058153113,0.022230634],"genre_scores_gemma":[0.19747004,0.00058379455,0.75924337,0.00044405274,0.00007318078,0.0025639385,0.03466426,0.0012607733,0.003696623],"study_design_codex":"design_other","study_design_gemma":"theoretical_or_conceptual","domain_scores_codex":[0.9779856,0.009104858,0.0040272176,0.003510428,0.0047534243,0.00061848375],"domain_scores_gemma":[0.9177834,0.04165714,0.0075557386,0.01622535,0.015351724,0.0014266885],"candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.02702418,0.0009206197,0.0011962085,0.022892518,0.0038330078,0.00893464,0.0023445282,0.0011770884,0.0031381024],"category_scores_gemma":[0.06588727,0.001117426,0.0009540666,0.02172525,0.0019914925,0.010880614,0.009095831,0.002061008,0.0016997805],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":null,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_system_candidate":false,"about_ca_system_consensus":false,"study_design_scores_codex":[0.0003087536,0.00059580605,0.11406042,0.0032673778,0.00030591091,0.002311643,0.091193624,0.0071841436,0.024984267,0.09389028,0.030766545,0.6311313],"study_design_scores_gemma":[0.00017060731,0.00027731928,0.082400955,0.0022581299,0.000379736,0.003076494,0.09938381,0.048744783,0.04705425,0.12249271,0.59322083,0.000540409],"about_ca_topic_score_codex":0.013581217,"about_ca_topic_score_gemma":0.02901388,"teacher_disagreement_score":0.02702418,"about_ca_system_score_codex":0.0029675993,"about_ca_system_score_gemma":0.009662352,"threshold_uncertainty_score":0.14291924},"labels":[],"label_agreement":null}]}