{"id":"W2062707524","doi":"10.1371/journal.pone.0115892","title":"Machine Learning for Biomedical Literature Triage","year":2014,"lang":"en","type":"article","venue":"PLoS ONE","topic":"Biomedical Text Mining and Ontologies","field":"Biochemistry, Genetics and Molecular Biology","cited_by":40,"is_retracted":false,"has_abstract":true,"ca_institutions":"Concordia University","funders":"Genome Alberta; Genome Canada","keywords":"Triage; Computer science; Machine learning; Artificial intelligence; Naive Bayes classifier; Support vector machine; Task (project management); Set (abstract data type); Domain (mathematical analysis); Logistic regression; Data mining; Medicine; Engineering; Mathematics","routes":{"ca_aff":true,"ca_fund":true,"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":[0.007336524,0.0009326569,0.00131799,0.005587585,0.001262769,0.002581075,0.001967076,0.001647997,0.004432319],"category_scores_gemma":[0.03992176,0.0004455683,0.001022063,0.005157851,0.0007746004,0.003584221,0.001983034,0.002212302,0.005295652],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001128456,"about_ca_system_score_gemma":0.002829461,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002394414,"about_ca_topic_score_gemma":0.002593185,"domain_scores_codex":[0.9940318,0.002177692,0.0006742,0.0009991054,0.001935557,0.0001816022],"domain_scores_gemma":[0.9749497,0.01577817,0.001735514,0.003828391,0.003226686,0.0004814923],"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":[0.0002804521,0.0002386178,0.004268201,0.0006805105,0.0001330463,0.0001853562,0.0002731381,0.01912241,0.00472292,0.01038897,0.0271538,0.9325526],"study_design_scores_gemma":[0.0001251894,0.000204323,0.0043241,0.0004300515,0.0001185912,0.0007878842,0.0003122945,0.7737452,0.01602718,0.1394759,0.06432862,0.0001207653],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.01193154,0.003380655,0.954833,0.003106641,0.0003668513,0.0003854751,0.001949773,0.02130665,0.002739386],"genre_scores_gemma":[0.1256577,0.001194149,0.866405,0.0006455537,0.0003278855,0.0004303993,0.003588404,0.0003806597,0.001370282],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.007336524,"threshold_uncertainty_score":0.0387997,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02749277960564432,"score_gpt":0.2516060964065917,"score_spread":0.2241133168009473,"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."}}