{"id":"W3161691007","doi":"10.1109/icse-companion52605.2021.00111","title":"An Open Dataset for Onboarding new Contributors: Empirical Study of OpenStack Ecosystem","year":2021,"lang":"en","type":"article","venue":"","topic":"Advanced Malware Detection Techniques","field":"Computer Science","cited_by":1,"is_retracted":false,"has_abstract":true,"ca_institutions":"Polytechnique Montréal; Queen's University","funders":"","keywords":"Onboarding; Computer science; Codebase; Empirical research; Data science; Software; Operating system","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":[0.0004343618,0.0001193057,0.000332397,0.00005956231,0.0001152739,0.0003265609,0.001728216,0.00004544996,0.00002996132],"category_scores_gemma":[0.0001476316,0.0001093548,0.00003003187,0.0003522065,0.000006615288,0.001417784,0.001137422,0.00006904046,0.000004899421],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00005785052,"about_ca_system_score_gemma":0.0001510986,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0001424601,"about_ca_topic_score_gemma":0.0008053317,"domain_scores_codex":[0.9985532,0.0001278126,0.0003710806,0.0005569455,0.0001904453,0.0002005374],"domain_scores_gemma":[0.9981377,0.0001644255,0.0001331565,0.001190517,0.0002382882,0.0001359233],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"not_applicable","study_design_gemma":"bench_or_experimental","study_design_scores_codex":[0.0005503684,0.009673451,0.007220101,0.0002854634,0.0006641233,0.0004409968,0.0051929,0.001996497,0.04644338,0.06263847,0.6042622,0.260632],"study_design_scores_gemma":[0.006097287,0.006089616,0.0007640535,0.00008059535,0.00004823168,0.00009447141,0.002680983,0.02709469,0.5155739,0.00594796,0.4346477,0.0008805293],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.01689739,0.000009896225,0.9807687,0.0002291285,0.0001547729,0.001237864,0.0003488834,0.0001958162,0.0001575877],"genre_scores_gemma":[0.678432,0.000001418682,0.3207956,0.000310384,0.00005238938,0.00009638799,0.0001588375,0.00001102479,0.0001419704],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.6615345,"threshold_uncertainty_score":0.4459361,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.08138962739021431,"score_gpt":0.4178714145246835,"score_spread":0.3364817871344691,"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."}}