{"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":"metacan-v3-hybrid-931329e0061c","candidate_categories":["metaresearch","open_science"],"consensus_categories":[],"category_scores_codex":[0.001487168,0.0006802009,0.0004278712,0.005155484,0.001060539,0.001078998,0.0009866379,0.001048287,0.00451387],"category_scores_gemma":[0.009330817,0.000194135,0.0004028086,0.005018759,0.0005788935,0.001722053,0.00215707,0.000879711,0.004566018],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.000821112,"about_ca_system_score_gemma":0.001183048,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.006015895,"about_ca_topic_score_gemma":0.01224526,"domain_scores_codex":[0.997486,0.000487483,0.0002758781,0.0005194213,0.0009619482,0.0002692577],"domain_scores_gemma":[0.9903424,0.003493289,0.001300968,0.001802347,0.002258352,0.0008026099],"domain_codex":null,"domain_gemma":"reproducibility","domain_candidate":"reproducibility","domain_consensus":null,"study_design_codex":"not_applicable","study_design_gemma":"observational","study_design_scores_codex":[0.0006601422,0.001319276,0.1342504,0.00219123,0.0001043345,0.001121311,0.005113134,0.002145167,0.004712526,0.003996181,0.7607788,0.08360749],"study_design_scores_gemma":[0.0001775998,0.000338185,0.3640398,0.0006396468,0.00005743397,0.001196785,0.007727543,0.008292614,0.006116433,0.003236529,0.6079969,0.0001804115],"study_design_candidate":"observational","study_design_consensus":null,"genre_codex":"dataset","genre_gemma":"empirical","genre_scores_codex":[0.2196759,0.0005273985,0.002873393,0.0006696154,0.0001672359,0.0004605108,0.7656897,0.002584993,0.007351293],"genre_scores_gemma":[0.102475,0.0001948468,0.006416053,0.0001673337,0.00007137651,0.001092504,0.8864977,0.0003185094,0.002766689],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.9990134,"threshold_uncertainty_score":0.01510036,"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."}}