{"id":"W4312064948","doi":"10.1111/hir.12471","title":"Application of text mining to the development and validation of a geographic search filter to facilitate evidence retrieval in Ovid <scp>MEDLINE</scp>: An example from the United States","year":2022,"lang":"en","type":"article","venue":"Health Information & Libraries Journal","topic":"Biomedical Text Mining and Ontologies","field":"Biochemistry, Genetics and Molecular Biology","cited_by":8,"is_retracted":false,"has_abstract":true,"ca_institutions":"Broadcom (Canada); Vancouver Coastal Health","funders":"","keywords":"MEDLINE; Filter (signal processing); Information retrieval; Computer science; Set (abstract data type); Vocabulary; Identification (biology); Controlled vocabulary; Data science; Data mining; Political science","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":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.00157777,0.0000682582,0.000108178,0.0001486088,0.0002871494,0.0000554186,0.0002376116,0.00003574797,0.000006321248],"category_scores_gemma":[0.0003302167,0.00004615097,0.00001629332,0.0004178787,0.00007299131,0.00005210541,0.000181696,0.0001565297,6.425557e-7],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00001813653,"about_ca_system_score_gemma":0.0002716458,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0005431155,"about_ca_topic_score_gemma":0.00004512143,"domain_scores_codex":[0.9986721,0.0002428696,0.0005523003,0.00009079768,0.0002818454,0.0001600598],"domain_scores_gemma":[0.9991543,0.0002579385,0.0002378618,0.0001547141,0.00009893009,0.00009622668],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"qualitative","study_design_gemma":"not_applicable","study_design_scores_codex":[0.001274759,0.0001227498,0.1463277,0.0002703909,0.0001056382,5.317382e-7,0.4977931,0.02838325,0.004358978,0.0002118574,0.01135572,0.3097953],"study_design_scores_gemma":[0.001096872,0.001897431,0.2611225,0.0001648025,0.000009589129,0.0000243875,0.1255329,0.007758759,0.02097516,0.0002180347,0.5810537,0.0001457517],"study_design_candidate":"qualitative","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9706344,0.0003436368,0.02536768,0.003316766,0.00003938759,0.0002466666,0.00004170977,0.000004848672,0.000004896853],"genre_scores_gemma":[0.9898031,0.0001518938,0.00748735,0.002066904,0.00003033003,0.00002752446,0.0004207255,0.000003587529,0.000008566479],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.569698,"threshold_uncertainty_score":0.220855,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1059803544471308,"score_gpt":0.3062490695496477,"score_spread":0.2002687151025169,"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."}}