{"id":"W1991705476","doi":"10.5555/2820518.2820577","title":"Mining StackOverflow to filter out off-topic IRC discussion","year":2015,"lang":"en","type":"article","venue":"Mining Software Repositories","topic":"Spam and Phishing Detection","field":"Computer Science","cited_by":25,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Alberta","funders":"","keywords":"Computer science; Exploit; Filter (signal processing); Set (abstract data type); The Internet; World Wide Web; Online discussion; Stack (abstract data type); Information retrieval; Open source; Data science; Programming language; Computer security","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":[0.003991825,0.00184665,0.001035227,0.01624718,0.001732869,0.002135377,0.001532695,0.002040391,0.001851881],"category_scores_gemma":[0.01750966,0.0004173713,0.001056247,0.005307382,0.0006421164,0.004592431,0.001686393,0.001310107,0.002714561],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001256926,"about_ca_system_score_gemma":0.002339584,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.004492614,"about_ca_topic_score_gemma":0.005524511,"domain_scores_codex":[0.996512,0.0007411117,0.0003211421,0.0008211769,0.00115789,0.0004467304],"domain_scores_gemma":[0.9820651,0.008163818,0.003370097,0.001161324,0.004396265,0.0008433696],"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":[0.0007740524,0.00136544,0.297419,0.001736716,0.0002662425,0.001766657,0.005359572,0.007168677,0.01896823,0.004375595,0.0593522,0.6014476],"study_design_scores_gemma":[0.0001509958,0.0009302601,0.2025099,0.0007649768,0.0006147137,0.002182121,0.008833152,0.5928681,0.06406649,0.0189973,0.107806,0.0002760314],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.8579358,0.003402831,0.1009326,0.001609072,0.0004854227,0.001007851,0.0116156,0.01038615,0.01262479],"genre_scores_gemma":[0.8717272,0.0009907193,0.07982119,0.0003435308,0.0005822,0.0008488946,0.03450322,0.000665212,0.01051787],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.01624718,"threshold_uncertainty_score":0.02111107,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03033946629969239,"score_gpt":0.2605105637548678,"score_spread":0.2301710974551754,"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."}}