{"id":"W2022780779","doi":"10.5539/cis.v4n6p48","title":"Semi-Automatic Labeling of Training Data Sets in Text Classification","year":2011,"lang":"en","type":"article","venue":"Computer and Information Science","topic":"Text and Document Classification Technologies","field":"Computer Science","cited_by":4,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"","keywords":"Carelessness; Computer science; Set (abstract data type); Training set; Process (computing); Reduction (mathematics); Data set; Data mining; Data reduction; Information retrieval; Artificial intelligence; Labeled data","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":[0.00560216,0.001027951,0.001314238,0.003664161,0.00187382,0.002233101,0.001950091,0.001516658,0.002199344],"category_scores_gemma":[0.0207616,0.0005364888,0.001059624,0.003105297,0.001152309,0.002822597,0.001643903,0.00165073,0.003843426],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0008803508,"about_ca_system_score_gemma":0.001968809,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002550632,"about_ca_topic_score_gemma":0.004965837,"domain_scores_codex":[0.9895025,0.005417819,0.001111236,0.001482291,0.002134655,0.0003513987],"domain_scores_gemma":[0.9604205,0.02297424,0.002364916,0.007279313,0.006533419,0.0004275218],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"bench_or_experimental","study_design_scores_codex":[0.000955034,0.001291876,0.008738826,0.001214474,0.0001175544,0.000267664,0.001223853,0.01757359,0.05996692,0.003528795,0.01424155,0.8908798],"study_design_scores_gemma":[0.0001870272,0.001122775,0.02525758,0.000509035,0.0002455651,0.001546622,0.001774839,0.6676418,0.2321016,0.01854468,0.05078263,0.0002858389],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.1367487,0.0008582422,0.8442156,0.0004311882,0.0002217261,0.001534148,0.002205369,0.008449119,0.005335876],"genre_scores_gemma":[0.2150701,0.0002923544,0.7726853,0.0001982092,0.00008346246,0.001290318,0.007597789,0.0003262413,0.002456168],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.00560216,"threshold_uncertainty_score":0.02962744,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1389177259685581,"score_gpt":0.3032114196743241,"score_spread":0.164293693705766,"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."}}