{"id":"W2987134874","doi":"10.48550/arxiv.1911.01217","title":"Detect Toxic Content to Improve Online Conversations","year":2019,"lang":"en","type":"preprint","venue":"arXiv (Cornell University)","topic":"Sentiment Analysis and Opinion Mining","field":"Computer Science","cited_by":3,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Ottawa","funders":"","keywords":"Python (programming language); Computer science; Naive Bayes classifier; Artificial intelligence; Support vector machine; Social media; Machine learning; Resampling; Natural language processing; Information retrieval; Deep learning; Content (measure theory); 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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0009492673,0.001121368,0.0005992049,0.001688948,0.0005055409,0.001534713,0.0006216994,0.001244644,0.003384639],"category_scores_gemma":[0.005672659,0.0002638761,0.0007475295,0.0007692585,0.0002642625,0.002478989,0.0009133714,0.001245412,0.003975586],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0006697892,"about_ca_system_score_gemma":0.0005700695,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002960759,"about_ca_topic_score_gemma":0.004889782,"domain_scores_codex":[0.9993413,0.0002015083,0.00003374586,0.000173288,0.0001536704,0.00009644679],"domain_scores_gemma":[0.9976949,0.001102525,0.0003168198,0.0001582191,0.0005752873,0.0001522409],"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":[0.001344336,0.00132826,0.1612586,0.000837516,0.000353936,0.0005857715,0.001817265,0.03816934,0.05247122,0.005445584,0.04861533,0.6877729],"study_design_scores_gemma":[0.00003305626,0.0004689407,0.04025571,0.0001408442,0.0001865655,0.0003748254,0.001472043,0.8896689,0.02895965,0.01361425,0.02475732,0.00006789745],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"methods","genre_scores_codex":[0.5414622,0.004352048,0.3839046,0.00527943,0.001146198,0.0006370051,0.01267478,0.01379049,0.03675332],"genre_scores_gemma":[0.9204743,0.0007909955,0.06121995,0.0005968756,0.0005612273,0.0001503686,0.007029221,0.000312691,0.008864405],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.003384639,"threshold_uncertainty_score":0.0113228,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1210382773475492,"score_gpt":0.2110812787755628,"score_spread":0.0900430014280136,"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."}}