{"id":"W4312019575","doi":"10.2196/40102","title":"Comparison of Methods for Estimating Temporal Topic Models From Primary Care Clinical Text Data: Retrospective Closed Cohort Study","year":2022,"lang":"en","type":"article","venue":"JMIR Medical Informatics","topic":"Machine Learning in Healthcare","field":"Computer Science","cited_by":21,"is_retracted":false,"has_abstract":true,"ca_institutions":"Public Health Ontario; University of Toronto","funders":"Canadian Institutes of Health Research; Heart and Stroke Foundation of Canada","keywords":"Latent Dirichlet allocation; Topic model; Computer science; Artificial intelligence; Latent class model; Machine learning","routes":{"ca_aff":true,"ca_fund":true,"ca_venue":true,"about_ca":true,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.005042577,0.000203837,0.0009411967,0.0001033287,0.0003483496,0.00006633996,0.003046499,0.0001518902,0.00004928332],"category_scores_gemma":[0.001121691,0.0001871549,0.0001116365,0.0003762823,0.0001061824,0.0006108865,0.003383715,0.00142708,0.000002308904],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0002490767,"about_ca_system_score_gemma":0.0007141469,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0002510228,"about_ca_topic_score_gemma":0.00001634578,"domain_scores_codex":[0.9946899,0.0009630371,0.002206094,0.0004151475,0.001383701,0.0003421641],"domain_scores_gemma":[0.9949954,0.001842119,0.0009694444,0.001692313,0.0002398027,0.0002609108],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.00003196841,0.0004946704,0.6923988,0.0004174208,0.0001013074,0.000003027314,0.034647,0.00218977,2.04306e-7,0.001047906,0.001716006,0.2669519],"study_design_scores_gemma":[0.0009703101,0.000821125,0.06807726,0.00004507972,0.00002694201,0.00000252889,0.005271258,0.9227406,9.783329e-7,0.0008059737,0.001067342,0.0001705858],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.1110608,0.0001198928,0.8845111,0.0002474578,0.001154845,0.002024604,0.0001031435,0.0001674044,0.000610832],"genre_scores_gemma":[0.3822335,0.000001501123,0.6166411,0.0003813839,0.0001381784,0.0002058091,0.0003790942,0.0000111007,0.000008319493],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.9205508,"threshold_uncertainty_score":0.7631955,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1450194641126829,"score_gpt":0.5170725829912691,"score_spread":0.3720531188785862,"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."}}