{"id":"W3190523902","doi":"10.1093/jamia/ocab124","title":"Application of natural language processing techniques to identify off-label drug usage from various online health communities","year":2021,"lang":"en","type":"article","venue":"Journal of the American Medical Informatics Association","topic":"Pharmaceutical studies and practices","field":"Medicine","cited_by":14,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Bristol-Myers Squibb Canada","keywords":"Drug; Spelling; Recall; Computer science; Multi-label classification; Off-label use; Parsing; Variety (cybernetics); Medicine; Natural language processing; Artificial intelligence; Psychology; Pharmacology","routes":{"ca_aff":false,"ca_fund":true,"ca_venue":false,"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":[0.003973038,0.0009890482,0.0004433979,0.006650934,0.0007301408,0.001340761,0.0006745045,0.0008191183,0.001704545],"category_scores_gemma":[0.01467534,0.0002917158,0.0007243697,0.002527339,0.0005241962,0.001939545,0.001218628,0.001099667,0.0008972893],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001099233,"about_ca_system_score_gemma":0.00168428,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.003456457,"about_ca_topic_score_gemma":0.005165472,"domain_scores_codex":[0.9967747,0.001306915,0.0004430529,0.0006601192,0.0006679298,0.0001472288],"domain_scores_gemma":[0.9655181,0.02611717,0.003763528,0.001086713,0.003140333,0.0003741121],"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":[0.001902601,0.001812446,0.09564111,0.003775195,0.0003090895,0.006008938,0.01053531,0.01430621,0.1063521,0.005329563,0.01794067,0.7360867],"study_design_scores_gemma":[0.0003493678,0.001471399,0.1773977,0.0008084511,0.0004930203,0.005273478,0.01012528,0.6073098,0.1087373,0.02742827,0.06013985,0.0004661035],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.5723936,0.0009480693,0.3806381,0.002125973,0.0002531488,0.003229043,0.02593357,0.007918966,0.00655953],"genre_scores_gemma":[0.5290544,0.0002641712,0.4499174,0.0003687916,0.0001267963,0.00142988,0.01727057,0.00014648,0.001421499],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.006650934,"threshold_uncertainty_score":0.02101165,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02667827383308776,"score_gpt":0.4245203176132349,"score_spread":0.3978420437801472,"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."}}