{"id":"W2077469553","doi":"10.3366/cor.2013.0032","title":"Challenges in cross-linguistic corpus-assisted discourse studies","year":2013,"lang":"en","type":"article","venue":"Corpora","topic":"Discourse Analysis in Language Studies","field":"Arts and Humanities","cited_by":60,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"","keywords":"Corpus linguistics; Linguistics; Focus (optics); Population; Contrastive linguistics; Sociology; Applied linguistics; Computer science; Philosophy","routes":{"ca_aff":false,"ca_fund":false,"ca_venue":false,"about_ca":true,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":["metaresearch"],"consensus_categories":[],"category_scores_codex":[0.4117099,0.001975871,0.003550462,0.03112472,0.02023621,0.03623689,0.01889822,0.009619286,0.01039582],"category_scores_gemma":[0.5745197,0.003212329,0.001259767,0.03175743,0.02574766,0.04698267,0.04933357,0.009424245,0.003786574],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.01383978,"about_ca_system_score_gemma":0.02250857,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01651374,"about_ca_topic_score_gemma":0.01638891,"domain_scores_codex":[0.4409367,0.4801955,0.02323239,0.0190866,0.03397805,0.00257078],"domain_scores_gemma":[0.1779298,0.6698792,0.01490087,0.06952594,0.06475298,0.003011228],"domain_codex":"methods","domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"qualitative","study_design_gemma":"theoretical_or_conceptual","study_design_scores_codex":[0.0002780728,0.0004877788,0.009779578,0.007594514,0.0005787511,0.00138188,0.3881352,0.002469599,0.002771949,0.2734862,0.0238877,0.2891489],"study_design_scores_gemma":[0.000192296,0.0001468118,0.008235915,0.0129167,0.0001668191,0.001307325,0.3078586,0.008631598,0.004146076,0.2297472,0.4263034,0.0003472566],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.1059377,0.05395466,0.5663074,0.1050988,0.005574013,0.009988272,0.006332242,0.001837263,0.1449697],"genre_scores_gemma":[0.4293774,0.009896004,0.5085256,0.008797354,0.001716607,0.02489738,0.005216195,0.00162959,0.009943843],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.4117099,"threshold_uncertainty_score":0.7254664,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.16594680154034,"score_gpt":0.363258996867498,"score_spread":0.1973121953271579,"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."}}