{"id":"W6983105518","doi":"","title":"Lemmatising Treebanks. Corpus Annotation with Knowledge Bases","year":2018,"lang":"en","type":"article","venue":"RIUR (Universidad de La Rioja)","topic":"Preterm Birth and Chorioamnionitis","field":"Medicine","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"","keywords":"Treebank; Annotation; Lemma (botany); Parsing; Set (abstract data type); Corpus linguistics; Text corpus","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":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0001193528,0.0001416475,0.000189098,0.0001481478,0.0001991525,0.00004443117,0.00008826273,0.000111807,0.0003463484],"category_scores_gemma":[0.00004748453,0.0001320747,0.00005629194,0.0003417992,0.0002770925,0.0002414808,0.00002822923,0.0001637529,0.0001599905],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001351873,"about_ca_system_score_gemma":0.0002434373,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0001175642,"about_ca_topic_score_gemma":0.000216045,"domain_scores_codex":[0.9991564,0.00007947403,0.00009802696,0.0002236408,0.000165382,0.0002770929],"domain_scores_gemma":[0.9992061,0.0001138842,0.0000729616,0.0002343355,0.0001742028,0.000198462],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"observational","study_design_scores_codex":[0.004131644,0.0008933679,0.2392303,0.000697611,0.0006805226,0.003642792,0.02505388,0.00001279657,0.04205497,0.0279022,0.05649137,0.5992085],"study_design_scores_gemma":[0.01132517,0.003086575,0.7504794,0.001165353,0.0005900356,0.002156631,0.002928161,0.004207767,0.01736756,0.001019877,0.20484,0.0008335069],"study_design_candidate":"observational","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.8558066,0.0002192287,0.002953586,0.0003941819,0.00009875591,0.0001572067,0.0000146478,0.0001510187,0.1402047],"genre_scores_gemma":[0.9926964,0.0000463529,0.00303367,0.000227411,0.0002806881,0.000002836395,0.00003463078,0.00002454622,0.003653448],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.598375,"threshold_uncertainty_score":0.5385849,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.00904552662072357,"score_gpt":0.2657146131267278,"score_spread":0.2566690865060042,"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."}}