{"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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.001727313,0.0005999577,0.0004757668,0.0007560675,0.0002955302,0.0008095186,0.0005996384,0.0005294825,0.00307666],"category_scores_gemma":[0.006318429,0.0001678512,0.0002603517,0.0004179789,0.0002149736,0.001215962,0.0008776243,0.0006184797,0.0009598836],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0005204431,"about_ca_system_score_gemma":0.0004248036,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.003281021,"about_ca_topic_score_gemma":0.004698254,"domain_scores_codex":[0.9989942,0.0003936054,0.00008677573,0.0002416278,0.0002196549,0.00006418909],"domain_scores_gemma":[0.9969832,0.001290392,0.0001992461,0.0003534463,0.00106239,0.0001113694],"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":[0.000856516,0.0007615199,0.01497685,0.0002829027,0.0001365356,0.0003618718,0.0003668208,0.03957462,0.05595846,0.0008169355,0.005015559,0.8808915],"study_design_scores_gemma":[0.0000468855,0.0003766963,0.01378827,0.00002541867,0.00003811423,0.0001306478,0.0002227992,0.936131,0.04696518,0.0005771233,0.001661824,0.00003604909],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.85569,0.0004251754,0.1313166,0.0002155936,0.000126466,0.0002188127,0.0007683676,0.006594815,0.004644107],"genre_scores_gemma":[0.9523484,0.00008083415,0.04262829,0.00006455019,0.00001385547,0.00009429042,0.001001994,0.00007833339,0.003689525],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.003281021,"threshold_uncertainty_score":0.01029247,"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."}}