{"id":"W4230896626","doi":"10.1109/msr.2015.54","title":"Mining StackOverflow to Filter Out Off-Topic IRC Discussion","year":2015,"lang":"en","type":"article","venue":"","topic":"Spam and Phishing Detection","field":"Computer Science","cited_by":31,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Alberta","funders":"Alberta Innovates - Technology Futures","keywords":"Computer science; Filter (signal processing); Exploit; Set (abstract data type); The Internet; World Wide Web; Stack (abstract data type); Open source; Information retrieval; Online discussion; Data science; Programming language; Computer security","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":[0.003120326,0.001943645,0.0009470297,0.01152644,0.001716794,0.002015029,0.001350685,0.002000113,0.00258925],"category_scores_gemma":[0.01198185,0.0003812085,0.001072685,0.00335658,0.0005645884,0.004173049,0.001363834,0.001411827,0.003205871],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001290634,"about_ca_system_score_gemma":0.002257102,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.005316803,"about_ca_topic_score_gemma":0.006711822,"domain_scores_codex":[0.9974011,0.0005360637,0.0002041302,0.0006672217,0.0007981258,0.0003933366],"domain_scores_gemma":[0.988473,0.005557779,0.00185623,0.0007079029,0.002750508,0.0006545022],"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":[0.001048834,0.001620278,0.2547584,0.00167039,0.0002843461,0.001823824,0.00438161,0.01130017,0.02288624,0.005350298,0.0805505,0.6143252],"study_design_scores_gemma":[0.0001366733,0.0007538694,0.1225063,0.0005102076,0.0004568992,0.001313803,0.005685449,0.7028596,0.05103074,0.01682932,0.09770571,0.0002113957],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.8313584,0.003915702,0.1158584,0.002126168,0.0007895018,0.001129392,0.01476723,0.01259317,0.01746197],"genre_scores_gemma":[0.8688035,0.0009621918,0.07487389,0.0003766808,0.0007617025,0.0007673998,0.03988659,0.0006552173,0.01291287],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.01152644,"threshold_uncertainty_score":0.01650208,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.04908666673615086,"score_gpt":0.2695334409950015,"score_spread":0.2204467742588506,"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."}}