{"id":"W3158640552","doi":"10.2196/27976","title":"Public Discussion of Anthrax on Twitter: Using Machine Learning to Identify Relevant Topics and Events","year":2021,"lang":"en","type":"article","venue":"JMIR Public Health and Surveillance","topic":"Data-Driven Disease Surveillance","field":"Medicine","cited_by":3,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"","keywords":"Categorization; Social media; Relevance (law); Bacillus anthracis; Data science; Information retrieval; Event (particle physics); Computer science; Artificial intelligence; Proxy (statistics); Machine learning; Natural language processing; World Wide Web; Biology","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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.002301832,0.0004139265,0.0003668453,0.004822372,0.0007405953,0.002054031,0.0003912573,0.0006403441,0.001223869],"category_scores_gemma":[0.008196438,0.0001925496,0.0004975416,0.002696604,0.0003563037,0.001956374,0.0007402908,0.000713003,0.0007139869],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0009538279,"about_ca_system_score_gemma":0.0005274417,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.004677394,"about_ca_topic_score_gemma":0.006198891,"domain_scores_codex":[0.9987478,0.0005192513,0.0001086683,0.0002173019,0.0002792077,0.0001276777],"domain_scores_gemma":[0.9912078,0.00560579,0.001696549,0.0002956266,0.0008653274,0.0003288916],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"observational","study_design_scores_codex":[0.0004853441,0.0007070105,0.7901281,0.0003384856,0.0001333975,0.00026274,0.00499866,0.004026665,0.005214981,0.0006869724,0.003675707,0.189342],"study_design_scores_gemma":[0.00003105161,0.0005303076,0.8115101,0.0002049707,0.0001247648,0.0002780196,0.009343857,0.1635176,0.005571169,0.002655797,0.006132072,0.0001003573],"study_design_candidate":"observational","study_design_consensus":"observational","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9841779,0.0002683255,0.00724161,0.0007187586,0.00004825984,0.0002816915,0.001686961,0.0001941995,0.005382309],"genre_scores_gemma":[0.9859324,0.0001898546,0.01142984,0.00008123813,0.00007819413,0.0001829904,0.001277311,0.00001409986,0.000814209],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.004822372,"threshold_uncertainty_score":0.01217335,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.07141352321373394,"score_gpt":0.3733272626778466,"score_spread":0.3019137394641127,"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."}}