{"id":"W4284891255","doi":"10.1017/s1351324922000298","title":"Neural automated writing evaluation for Korean L2 writing","year":2022,"lang":"en","type":"article","venue":"Natural Language Engineering","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":12,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of British Columbia","funders":"National Research Foundation of Korea; National Research Foundation","keywords":"Computer science; Fluency; Parsing; Artificial intelligence; Natural language processing; Complement (music); Artificial neural network; Reliability (semiconductor); Linguistics","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":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.001108374,0.0002171628,0.0001918095,0.0002587484,0.0003552236,0.0002218204,0.0009007987,0.00005639517,0.00001941663],"category_scores_gemma":[0.0004326135,0.0002245338,0.0001055985,0.0006237628,0.000009390847,0.0005960084,0.0004566852,0.0005617711,0.000001321109],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0003412373,"about_ca_system_score_gemma":0.00005090825,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0000255485,"about_ca_topic_score_gemma":0.000001844968,"domain_scores_codex":[0.9980986,0.00006971094,0.000312992,0.0004371761,0.0006080955,0.0004734973],"domain_scores_gemma":[0.999097,0.0002181747,0.0001196973,0.0003599631,0.0001425117,0.00006259037],"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":[0.00002870201,0.00006746413,0.00009741817,0.0003281744,0.000074225,0.0001906239,0.005250131,0.04025569,0.2645046,0.01921583,0.001666892,0.6683203],"study_design_scores_gemma":[0.0003380151,0.00004411106,0.00006971575,0.00003145617,0.00001257398,0.000106583,0.0002112733,0.9869491,0.0116577,0.0001573684,0.0001228251,0.0002992755],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.4243174,0.06325542,0.4789334,0.001611586,0.003357625,0.002682338,0.00006777687,0.02552934,0.0002452019],"genre_scores_gemma":[0.7439691,5.73665e-7,0.2553976,0.0001594364,0.0001613441,0.0001901777,0.00006201126,0.00002895791,0.00003076543],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.9466934,"threshold_uncertainty_score":0.9156221,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.009429505732520903,"score_gpt":0.2826202360865886,"score_spread":0.2731907303540677,"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."}}