{"id":"W4225143092","doi":"10.1145/3533021","title":"A Machine Learning Approach for Automated Filling of Categorical Fields in Data Entry Forms","year":2022,"lang":"en","type":"article","venue":"ACM Transactions on Software Engineering and Methodology","topic":"Data Quality and Management","field":"Decision Sciences","cited_by":13,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Ottawa","funders":"","keywords":"Categorical variable; Computer science; Field (mathematics); Machine learning; Data mining; Artificial intelligence; Set (abstract data type); Data field; Software; 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":[0.005326908,0.00009919571,0.0003011914,0.0003548315,0.0001403948,0.00001883535,0.0007171597,0.00006303361,0.00005321924],"category_scores_gemma":[0.004852156,0.00008775925,0.00004868455,0.0004366402,0.00002259414,0.0001227743,0.0001244106,0.0003524248,5.061851e-7],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00002283164,"about_ca_system_score_gemma":0.00002005836,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00009879094,"about_ca_topic_score_gemma":0.00001662239,"domain_scores_codex":[0.9983908,0.0003678392,0.000404708,0.0003978476,0.0002475862,0.0001912311],"domain_scores_gemma":[0.9939123,0.005312463,0.00007821657,0.0006303177,0.00002462029,0.00004203281],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.00009696207,0.0001110462,0.0001256048,0.00005746103,0.00004495267,0.000002497497,0.0006565891,0.9154909,0.00003320824,0.0008473601,0.0002175657,0.08231585],"study_design_scores_gemma":[0.000773358,0.0002252648,0.0002998557,0.000005198756,0.00003990001,0.00002390135,0.0009009524,0.9624075,0.0001119036,0.002630606,0.03241669,0.0001649071],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.003516193,0.0001680727,0.9953588,0.0002294182,0.0002110895,0.0001754777,0.0001940604,0.0001372394,0.000009670572],"genre_scores_gemma":[0.2968627,0.00005559317,0.7024347,0.00009472036,0.0000108852,0.00009923197,0.0001829299,0.00001247918,0.0002467584],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.2933465,"threshold_uncertainty_score":0.5808835,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.3306833255399733,"score_gpt":0.4220707860804389,"score_spread":0.09138746054046559,"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."}}