{"id":"W4220759704","doi":"10.3390/app12062891","title":"Survey of BERT-Base Models for Scientific Text Classification: COVID-19 Case Study","year":2022,"lang":"en","type":"article","venue":"Applied Sciences","topic":"Topic Modeling","field":"Computer Science","cited_by":86,"is_retracted":false,"has_abstract":true,"ca_institutions":"Université de Moncton","funders":"","keywords":"Coronavirus disease 2019 (COVID-19); Computer science; Scientific literature; Pandemic; Data science; Context (archaeology); Task (project management); Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2); Domain (mathematical analysis); Artificial intelligence; History; Infectious disease (medical specialty); Medicine; Engineering; Disease; Mathematics","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":[0.007426017,0.00289112,0.001825805,0.009011217,0.001138426,0.002953183,0.003482593,0.002285483,0.002747431],"category_scores_gemma":[0.01838975,0.0008650042,0.002363108,0.006413464,0.0006440734,0.004123918,0.001531082,0.003009939,0.004355263],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.002439863,"about_ca_system_score_gemma":0.003589396,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.02006884,"about_ca_topic_score_gemma":0.02870511,"domain_scores_codex":[0.9967399,0.001401195,0.0004449753,0.0005424219,0.0006877406,0.0001837761],"domain_scores_gemma":[0.9820239,0.01348895,0.0005702202,0.001006395,0.002330374,0.0005801943],"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":[0.001219697,0.0009030568,0.02255844,0.003831805,0.001218856,0.0004990508,0.0005540116,0.1167036,0.004513932,0.007347565,0.1162103,0.7244398],"study_design_scores_gemma":[0.00008803258,0.0004986486,0.006279434,0.0006601429,0.0003263612,0.0004829695,0.0004187329,0.9132473,0.003793511,0.01314626,0.06093455,0.0001240417],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.171546,0.2091439,0.4808636,0.02320131,0.003813768,0.002032085,0.04835306,0.03210949,0.02893681],"genre_scores_gemma":[0.5421573,0.04975136,0.2730275,0.004327251,0.002803592,0.001218463,0.1082304,0.00145165,0.01703244],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.02006884,"threshold_uncertainty_score":0.039904,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.2673387486849687,"score_gpt":0.3575259476444219,"score_spread":0.09018719895945321,"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."}}