{"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":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0003327503,0.0002017715,0.0002040439,0.0001144925,0.0002670347,0.0005623002,0.0005618529,0.0001090604,0.000003116674],"category_scores_gemma":[0.000700165,0.0001507482,0.00006757408,0.0003088235,0.00002680925,0.0007322612,0.0003587127,0.0001194203,0.00004150812],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001344288,"about_ca_system_score_gemma":0.0001232514,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00004216571,"about_ca_topic_score_gemma":0.00003876085,"domain_scores_codex":[0.9982518,0.0000810835,0.0002856272,0.0005153838,0.0005213058,0.0003447824],"domain_scores_gemma":[0.9986305,0.0001380629,0.000109261,0.0006727214,0.0001867129,0.0002627298],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"not_applicable","study_design_scores_codex":[0.000126423,0.0001280468,0.2639435,0.00008479397,0.00009208696,0.0003069552,0.1773675,0.0008480599,0.001611196,0.0004434506,0.09591822,0.4591298],"study_design_scores_gemma":[0.002864822,0.003424965,0.08736383,0.00143079,0.0001322128,0.0004913708,0.007730105,0.008710708,0.06062886,0.003023971,0.8206048,0.003593536],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.4689188,0.0002594494,0.512066,0.001409069,0.01495855,0.0002086227,0.000003118439,0.001157434,0.001018932],"genre_scores_gemma":[0.7241033,0.000001940179,0.2672758,0.0002931031,0.001505115,0.00003983141,0.000005936,0.00003281481,0.006742161],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.7246866,"threshold_uncertainty_score":0.6147331,"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."}}