{"id":"W4366966678","doi":"10.1109/wi-iat55865.2022.00024","title":"NLI-based Filtering for Data Augmentation in Topic Classification","year":2022,"lang":"en","type":"article","venue":"","topic":"Topic Modeling","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Wilfrid Laurier University","funders":"Natural Sciences and Engineering Research Council of Canada","keywords":"Computer science; Leverage (statistics); Artificial intelligence; Machine learning; Task (project management); Filter (signal processing); Inference; Generalization; Encoder; Data mining; Language model; Natural language processing","routes":{"ca_aff":true,"ca_fund":true,"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":[0.01194671,0.001648759,0.001824958,0.002872764,0.001596245,0.002086335,0.00335963,0.002151439,0.003723889],"category_scores_gemma":[0.03330007,0.0008705565,0.00259521,0.002847891,0.001711416,0.004489403,0.002860563,0.004016907,0.003259085],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001396645,"about_ca_system_score_gemma":0.002684503,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.004078442,"about_ca_topic_score_gemma":0.009515719,"domain_scores_codex":[0.9948355,0.002308246,0.0003859095,0.001474064,0.0007370795,0.0002592493],"domain_scores_gemma":[0.9794945,0.01285564,0.0008441132,0.004807019,0.001702371,0.0002964928],"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":[0.0007634893,0.0007230651,0.01011064,0.0007879693,0.0002560185,0.0001901518,0.001268501,0.05610155,0.02349455,0.02086699,0.01690198,0.8685351],"study_design_scores_gemma":[0.0001183651,0.0002294526,0.003104805,0.0001360036,0.0001164307,0.0002799557,0.0002364359,0.9187123,0.02163171,0.03738146,0.01797632,0.000076744],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.007990923,0.0005124101,0.9857255,0.0003393675,0.0001064685,0.0002629712,0.0004713662,0.003848403,0.0007425762],"genre_scores_gemma":[0.1312507,0.0003560653,0.8602511,0.000561849,0.000256029,0.0009799246,0.004159986,0.0006096368,0.001574824],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.01194671,"threshold_uncertainty_score":0.06318098,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.2319445510874427,"score_gpt":0.3477702179223591,"score_spread":0.1158256668349164,"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."}}