{"id":"W3126565418","doi":"10.2196/21679","title":"Lexicon Development for COVID-19-related Concepts Using Open-source Word Embedding Sources: An Intrinsic and Extrinsic Evaluation","year":2021,"lang":"en","type":"article","venue":"JMIR Medical Informatics","topic":"Biomedical Text Mining and Ontologies","field":"Biochemistry, Genetics and Molecular Biology","cited_by":8,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"National Center for Advancing Translational Sciences; National Institutes of Health; Perelman School of Medicine, University of Pennsylvania; University of Pennsylvania","keywords":"Coronavirus disease 2019 (COVID-19); Lexicon; Computer science; Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2); Word embedding; 2019-20 coronavirus outbreak; Word (group theory); Open source; Natural language processing; Artificial intelligence; Embedding; Linguistics; Medicine; Virology; Software; Pathology; Programming language","routes":{"ca_aff":false,"ca_fund":false,"ca_venue":true,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.01541613,0.002448847,0.001026203,0.00981826,0.001319795,0.005133424,0.00165183,0.00183592,0.00598075],"category_scores_gemma":[0.08242375,0.0006212975,0.002165012,0.004791855,0.001106909,0.009554804,0.008212369,0.002141083,0.003573618],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001846107,"about_ca_system_score_gemma":0.003098901,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.004079839,"about_ca_topic_score_gemma":0.005620226,"domain_scores_codex":[0.98643,0.005748508,0.002474438,0.001778935,0.003156752,0.0004113297],"domain_scores_gemma":[0.9333894,0.04562182,0.002928961,0.005228843,0.0118003,0.001030765],"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":[0.002525783,0.001435186,0.06646337,0.008420877,0.001023564,0.001735586,0.007625359,0.01201962,0.03029931,0.01545555,0.08656084,0.7664351],"study_design_scores_gemma":[0.001288926,0.002198549,0.0592045,0.003272075,0.001680445,0.004318805,0.01338277,0.6219828,0.06410609,0.03941884,0.1883661,0.0007801239],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.4028084,0.004166462,0.4884005,0.003272541,0.00161023,0.006091319,0.04294391,0.03112238,0.01958431],"genre_scores_gemma":[0.3463165,0.001264101,0.5619928,0.0005357724,0.0001807593,0.003351929,0.08081139,0.002363301,0.003183458],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.01541613,"threshold_uncertainty_score":0.08152926,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.07589123149888627,"score_gpt":0.4143356395403869,"score_spread":0.3384444080415006,"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."}}