{"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":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0003464068,0.00006657164,0.0001138953,0.0001482756,0.0002723959,0.00004150037,0.000536904,0.000002988787,0.0002006208],"category_scores_gemma":[8.701331e-7,0.00006297943,0.00004151596,0.000697072,0.00006532915,0.00006932732,0.000376581,0.0001011077,0.000003731051],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00003752821,"about_ca_system_score_gemma":0.0000344287,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0001190981,"about_ca_topic_score_gemma":0.000002114069,"domain_scores_codex":[0.9988912,0.00004666595,0.0001563803,0.0003048333,0.0004646327,0.0001362688],"domain_scores_gemma":[0.9993872,0.0000279325,0.00009483797,0.0003899052,0.00007009403,0.0000300229],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"observational","study_design_scores_codex":[0.00001237388,0.001838194,0.8694375,0.000002565076,0.00003903527,0.000002098519,0.002403754,0.006329845,0.0007709352,0.02494631,0.002980737,0.09123665],"study_design_scores_gemma":[0.0003121873,0.0008307557,0.8122804,0.000004642922,0.00002064224,9.11561e-7,0.0004137069,0.1811357,0.0004080438,0.00325061,0.001169325,0.0001731439],"study_design_candidate":"observational","study_design_consensus":"observational","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9054469,0.000001622156,0.09315315,0.0001160791,0.000108017,0.0001409348,0.000001006092,0.00002678261,0.001005498],"genre_scores_gemma":[0.9975954,2.752385e-8,0.002139061,0.00003808706,0.0001102286,0.00003748411,0.000002011001,0.000003003112,0.00007466532],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.1748058,"threshold_uncertainty_score":0.2568226,"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."}}