{"id":"W2906789718","doi":"10.2196/12414","title":"Data Analysis and Visualization of Newspaper Articles on Thirdhand Smoke: A Topic Modeling Approach","year":2019,"lang":"en","type":"article","venue":"JMIR Medical Informatics","topic":"Smoking Behavior and Cessation","field":"Medicine","cited_by":20,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"National Social Science Fund of China","keywords":"Newspaper; Visualization; Data visualization; Computer science; Data science; Topic model; Information retrieval; Data mining; Advertising; Business","routes":{"ca_aff":false,"ca_fund":false,"ca_venue":true,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0003361772,0.00007562314,0.0002371795,0.0001703181,0.00002209413,0.00001771697,0.0001045293,0.0001264078,0.0001498424],"category_scores_gemma":[0.00007895824,0.00005571717,0.00003371944,0.0003667841,0.00003630234,0.0001806572,0.00007203835,0.0001273804,0.000002744701],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00001460419,"about_ca_system_score_gemma":0.00004745157,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0000229516,"about_ca_topic_score_gemma":0.00000254141,"domain_scores_codex":[0.9987394,0.00001577798,0.0003866058,0.00009499225,0.0006610196,0.000102173],"domain_scores_gemma":[0.9992924,0.00003338173,0.00009210616,0.0004216863,0.00005440258,0.0001060783],"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":[0.0001284348,0.0005363881,0.9466689,0.001000541,0.0004217751,0.000002097116,0.01125916,0.0006246366,0.0001178001,0.001973264,0.0002787773,0.03698828],"study_design_scores_gemma":[0.001022196,0.0001212579,0.05763103,0.0001072799,0.0004294173,0.000007884512,0.002056976,0.9382806,0.0001311988,0.00003295885,0.0001001494,0.0000790815],"study_design_candidate":"observational","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9780878,0.00001105059,0.0198294,0.00007377042,0.00003203327,0.0002112922,0.000006015478,0.00002576061,0.001722898],"genre_scores_gemma":[0.9985375,0.00003131117,0.0006475994,0.0003814467,0.0000297208,0.000005956902,0.0003095592,0.000005542611,0.00005140485],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.9376559,"threshold_uncertainty_score":0.227208,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.06729296493470174,"score_gpt":0.3616715974657648,"score_spread":0.294378632531063,"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."}}