{"id":"W2401290433","doi":"10.1145/2901739.2901770","title":"Mining duplicate questions in stack overflow","year":2016,"lang":"en","type":"article","venue":"","topic":"Software Engineering Research","field":"Computer Science","cited_by":123,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Saskatchewan","funders":"","keywords":"Computer science; Reputation; Recall rate; Information retrieval; Precision and recall; Data mining; Recall; Stack (abstract data type); Data science; World Wide Web; Artificial intelligence; Programming language","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":[0.008595061,0.001363714,0.001436585,0.01173679,0.00193515,0.002399173,0.002174758,0.003220636,0.001332804],"category_scores_gemma":[0.06134528,0.0006429168,0.001048387,0.004976668,0.001154082,0.005091816,0.003421395,0.00128531,0.0008816086],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001283797,"about_ca_system_score_gemma":0.002274,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.004195483,"about_ca_topic_score_gemma":0.003941019,"domain_scores_codex":[0.9859735,0.003572185,0.002075047,0.002479021,0.004996622,0.0009037326],"domain_scores_gemma":[0.9286872,0.04441668,0.009548705,0.004336383,0.01150736,0.001503792],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"observational","study_design_scores_codex":[0.001498056,0.0008691598,0.3928545,0.003702325,0.000401791,0.01016475,0.02172257,0.008825946,0.04454207,0.008522521,0.02850898,0.4783873],"study_design_scores_gemma":[0.0002069782,0.001344915,0.3163691,0.00136924,0.001128721,0.02456906,0.02361065,0.3322924,0.1414984,0.03122321,0.125875,0.0005123265],"study_design_candidate":"observational","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.8769137,0.003130715,0.1051432,0.0008719733,0.0002148339,0.0008392282,0.005495946,0.003886516,0.003503978],"genre_scores_gemma":[0.848065,0.0008017835,0.1308919,0.0006605784,0.0002622495,0.0005589442,0.01370561,0.000400705,0.004653338],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.01173679,"threshold_uncertainty_score":0.04545557,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02278426483884391,"score_gpt":0.2797278117024878,"score_spread":0.2569435468636439,"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."}}