{"id":"W4379652851","doi":"10.1136/annrheumdis-2023-eular.6231","title":"AB1767-HPR DOCUMENT SEARCH IN LARGE RHEUMATOLOGY DATABASES: ADVANCED KEYWORD QUERIES TO SELECT HOMOGENEOUS PHENOTYPES","year":2023,"lang":"en","type":"preprint","venue":"Annals of the Rheumatic Diseases","topic":"Biomedical Text Mining and Ontologies","field":"Biochemistry, Genetics and Molecular Biology","cited_by":0,"is_retracted":false,"has_abstract":false,"ca_institutions":"McGill University; University of British Columbia","funders":"","keywords":"Computer science; Homogeneous; Keyword search; Information retrieval; Database; World Wide Web; 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":[0.001561757,0.001479569,0.001039628,0.00442839,0.0005405587,0.002370589,0.001113267,0.001666363,0.03119303],"category_scores_gemma":[0.007959223,0.0005664085,0.001045747,0.003581569,0.0003167593,0.002161632,0.001582442,0.0006501426,0.01865955],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0006121257,"about_ca_system_score_gemma":0.0014405,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.005299026,"about_ca_topic_score_gemma":0.006465352,"domain_scores_codex":[0.9987721,0.0002459913,0.0001788219,0.0002929289,0.0004014168,0.0001087483],"domain_scores_gemma":[0.9978722,0.001139321,0.0001028745,0.0004410572,0.0003194701,0.0001249502],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"not_applicable","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.002666617,0.0005320487,0.006651891,0.003038118,0.0006226449,0.001182949,0.0005267669,0.007660562,0.02840741,0.0140187,0.6501917,0.2845007],"study_design_scores_gemma":[0.002973606,0.0004954276,0.02075375,0.0004601526,0.0005901742,0.002752931,0.001071969,0.2521627,0.05822075,0.04824222,0.6120584,0.0002179041],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"dataset","genre_gemma":"methods","genre_scores_codex":[0.07005133,0.003041069,0.1794692,0.002981494,0.0004957267,0.0007768263,0.4713764,0.2421216,0.02968619],"genre_scores_gemma":[0.1203654,0.001436572,0.2747858,0.0005674687,0.0002592558,0.000481161,0.5802231,0.009906476,0.01197479],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.03119303,"threshold_uncertainty_score":0.1043512,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.05282673366597263,"score_gpt":0.3598819778971527,"score_spread":0.3070552442311801,"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."}}