{"id":"W4388777264","doi":"10.2196/50150","title":"A Comparison of ChatGPT and Fine-Tuned Open Pre-Trained Transformers (OPT) Against Widely Used Sentiment Analysis Tools: Sentiment Analysis of COVID-19 Survey Data","year":2023,"lang":"en","type":"article","venue":"JMIR Mental Health","topic":"Sentiment Analysis and Opinion Mining","field":"Computer Science","cited_by":72,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"National Institute of Arthritis and Musculoskeletal and Skin Diseases; National Institutes of Health","keywords":"Sentiment analysis; Computer science; Social media; Context (archaeology); Data science; Sentence; Coronavirus disease 2019 (COVID-19); Artificial intelligence; Data mining; World Wide Web; Medicine","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":["metaepi_narrow"],"consensus_categories":[],"category_scores_codex":[0.004378299,0.0003538062,0.001832649,0.001752117,0.0003454496,0.000270481,0.001941118,0.00008324072,0.0001073148],"category_scores_gemma":[0.00004407974,0.0003343224,0.0004080957,0.009361333,0.0001186493,0.0008129223,0.00126869,0.0001341947,0.00000607072],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.000242475,"about_ca_system_score_gemma":0.0004548702,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.004004981,"about_ca_topic_score_gemma":0.006353661,"domain_scores_codex":[0.9946105,0.0007171453,0.001752363,0.001194805,0.001114005,0.0006111243],"domain_scores_gemma":[0.9964967,0.0003562392,0.001083571,0.001437974,0.00009095432,0.0005345668],"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":[0.0004354992,0.00211061,0.8690078,0.0005636929,0.02287325,0.000009071451,0.04540695,0.01002826,0.002075046,0.0003608391,0.01111507,0.03601392],"study_design_scores_gemma":[0.001914948,0.0002840067,0.2960072,0.00005461023,0.001026673,4.106006e-7,0.002671264,0.6960236,0.0008805546,0.000007302221,0.0007670714,0.0003623576],"study_design_candidate":"observational","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.954325,0.0004322616,0.03665111,0.004235926,0.0001796804,0.002092542,0.001874002,0.0001267339,0.00008280615],"genre_scores_gemma":[0.9895697,0.0001747011,0.002397644,0.000425411,0.00001134889,0.00004629756,0.007120927,0.00001671828,0.000237208],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.6859953,"threshold_uncertainty_score":0.9999109,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1909377796622881,"score_gpt":0.4657761489750528,"score_spread":0.2748383693127647,"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."}}