{"id":"W4224262247","doi":"10.1007/s42488-022-00066-6","title":"Classifying multi-level product categories using dynamic masking and transformer models","year":2022,"lang":"en","type":"article","venue":"Journal of Data Information and Management","topic":"Text and Document Classification Technologies","field":"Computer Science","cited_by":10,"is_retracted":false,"has_abstract":false,"ca_institutions":"Toronto Metropolitan University","funders":"","keywords":"Computer science; Categorization; Transformer; Generalizability theory; Machine learning; Artificial intelligence; Product (mathematics); Deep learning; Natural language processing; Engineering; 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":[0.000533114,0.00006901592,0.00009055549,0.0003060905,0.0002574137,0.0002718616,0.0005972887,0.0000112076,0.000003419072],"category_scores_gemma":[0.000006777918,0.00005987774,0.00001324944,0.0001848981,0.00002699241,0.007767906,0.0005978383,0.000115149,4.432483e-7],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00005056345,"about_ca_system_score_gemma":0.00002552545,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.000003988587,"about_ca_topic_score_gemma":0.000001674352,"domain_scores_codex":[0.9991726,0.00001640304,0.0003634675,0.00009465605,0.0002545564,0.00009829424],"domain_scores_gemma":[0.999346,0.000008105779,0.0003015316,0.0002778123,0.00003999152,0.00002658943],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.00001109777,0.00003134037,0.00005407081,0.0001206807,0.0000530226,0.000003486318,0.001704403,0.002881723,0.00006008167,0.06881963,0.0005932843,0.9256672],"study_design_scores_gemma":[0.0007201938,0.00004732531,0.001289546,0.0000193197,0.0000273017,0.00009829992,0.004671457,0.939689,0.00004648453,0.002549281,0.05070797,0.0001338607],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.002501555,0.0002679152,0.9950997,0.001615929,0.0001405181,0.0001475074,0.00001388259,0.00003361897,0.0001793871],"genre_scores_gemma":[0.7604071,0.001153277,0.2381303,0.0002118782,0.000006104005,0.000006063149,0.0000226063,0.000002909227,0.00005968188],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.9368072,"threshold_uncertainty_score":0.5631545,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1441194875508052,"score_gpt":0.3142199106093222,"score_spread":0.170100423058517,"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."}}