{"id":"W4399158753","doi":"10.1075/kl.00008.par","title":"Word segmentation granularity in Korean","year":2024,"lang":"en","type":"article","venue":"Korean Linguistics","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of British Columbia","funders":"","keywords":"Granularity; Segmentation; Natural language processing; Word (group theory); Computer science; Text segmentation; Linguistics; Artificial intelligence; Philosophy","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":[0.0005783,0.0003953405,0.0002272834,0.001857653,0.0006351416,0.001977321,0.0003901735,0.0003027297,0.005339504],"category_scores_gemma":[0.001975131,0.0003760624,0.0004418192,0.002280889,0.0008854687,0.00297527,0.001147608,0.0005377221,0.001537821],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0007623952,"about_ca_system_score_gemma":0.0007889948,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002892911,"about_ca_topic_score_gemma":0.003343519,"domain_scores_codex":[0.9993091,0.000118853,0.00009273439,0.0002635732,0.0001217786,0.00009410863],"domain_scores_gemma":[0.9985656,0.0004442648,0.0002443691,0.0002633563,0.0004054078,0.00007700018],"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":[0.001049042,0.0001275235,0.05508684,0.001108608,0.0001015743,0.001945295,0.01473166,0.01210824,0.1686548,0.1414365,0.00875942,0.5948905],"study_design_scores_gemma":[0.0001102594,0.0005201552,0.1392431,0.0007182343,0.00047126,0.003181495,0.01729959,0.121974,0.2115666,0.1888352,0.3156051,0.0004749657],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.7163243,0.002757384,0.2135493,0.0006481177,0.0001664881,0.0002254333,0.002424152,0.003017968,0.06088678],"genre_scores_gemma":[0.9280754,0.0004629034,0.0659352,0.0001387993,0.00001832682,0.00004625246,0.00144247,0.0002975755,0.003583089],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.005339504,"threshold_uncertainty_score":0.01786244,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01480372139418494,"score_gpt":0.2943217223496563,"score_spread":0.2795180009554714,"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."}}