{"id":"W2620887933","doi":"10.7717/peerj-cs.991","title":"A longitudinal study of topic classification on Twitter","year":2022,"lang":"en","type":"article","venue":"PeerJ Computer Science","topic":"Complex Network Analysis Techniques","field":"Physics and Astronomy","cited_by":8,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Toronto","funders":"","keywords":"Classifier (UML); Computer science; Artificial intelligence; Machine learning; Generalization; Social media; Feature (linguistics); Entertainment; Natural language processing; Information retrieval; 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":[0.004226204,0.0002902455,0.00046518,0.001479557,0.001248117,0.00190131,0.0005300531,0.0008909756,0.003130202],"category_scores_gemma":[0.02517298,0.0002857856,0.0004005517,0.001596528,0.000569954,0.003530642,0.00126054,0.001619475,0.001790775],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0006848221,"about_ca_system_score_gemma":0.0004403293,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.008646393,"about_ca_topic_score_gemma":0.01014198,"domain_scores_codex":[0.9983048,0.0007294237,0.00009760479,0.0003806542,0.000280286,0.0002072222],"domain_scores_gemma":[0.9832308,0.007519024,0.002909583,0.001926647,0.003116943,0.001296919],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"observational","study_design_scores_codex":[0.0002848463,0.0004261297,0.9436722,0.00006920162,0.0001111839,0.0002415904,0.006385053,0.0009445653,0.001317003,0.001790767,0.01174811,0.03300921],"study_design_scores_gemma":[0.00002418475,0.0004032109,0.9317375,0.0001042563,0.00008268657,0.0004109491,0.008865203,0.02869066,0.001332174,0.002585303,0.02567852,0.00008528872],"study_design_candidate":"observational","study_design_consensus":"observational","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9876344,0.0007090123,0.002782464,0.00206982,0.0001395449,0.00005976814,0.003197301,0.00009531268,0.003312416],"genre_scores_gemma":[0.9922417,0.0002261283,0.001026564,0.0002706873,0.0001642588,0.00007260294,0.003785302,0.0000547386,0.00215795],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.008646393,"threshold_uncertainty_score":0.02235061,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.0619292373093529,"score_gpt":0.3235566165200028,"score_spread":0.2616273792106499,"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."}}