{"id":"W4226024751","doi":"10.1017/s1351324922000134","title":"Real-world sentence boundary detection using multitask learning: A case study on French","year":2022,"lang":"en","type":"article","venue":"Natural Language Engineering","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":3,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of British Columbia","funders":"Hanbat National University","keywords":"Computer science; Sentence; Punctuation; Natural language processing; Task (project management); Artificial intelligence; Boundary (topology); Multi-task learning; Speech recognition","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":[0.002485326,0.0008659191,0.0005160004,0.001920444,0.001549806,0.001315851,0.001101493,0.001910853,0.001264634],"category_scores_gemma":[0.01215881,0.0001517594,0.0005616944,0.001707702,0.00075051,0.001325161,0.0007214783,0.0007789317,0.0007463995],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001269524,"about_ca_system_score_gemma":0.0007627204,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.03128308,"about_ca_topic_score_gemma":0.04338316,"domain_scores_codex":[0.9972603,0.001534842,0.0001538556,0.0004492524,0.0004378481,0.0001638569],"domain_scores_gemma":[0.9872562,0.008560916,0.0006162623,0.0009797586,0.002082011,0.000504921],"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":[0.003340783,0.003347955,0.1141322,0.003781487,0.0007039466,0.03303708,0.01725966,0.06382222,0.08793525,0.006566803,0.09650704,0.5695656],"study_design_scores_gemma":[0.0006159162,0.002433633,0.2399611,0.0004910034,0.0004445428,0.01229638,0.01976743,0.4120922,0.1236983,0.01337952,0.1742778,0.0005421676],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9574623,0.001284699,0.02884295,0.001378357,0.0001512362,0.0002353815,0.003811116,0.002708739,0.004125145],"genre_scores_gemma":[0.9515767,0.0002393936,0.03930082,0.000377801,0.000102889,0.0001506852,0.005928485,0.0001912287,0.002132082],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.03128308,"threshold_uncertainty_score":0.06220198,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01083990589589887,"score_gpt":0.2761478394611767,"score_spread":0.2653079335652778,"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."}}