{"id":"W3089977198","doi":"10.1145/3377812.3382165","title":"An empirical study of the first contributions of developers to open source projects on GitHub","year":2020,"lang":"en","type":"article","venue":"","topic":"Open Source Software Innovations","field":"Computer Science","cited_by":9,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Waterloo","funders":"","keywords":"Java; Open source; Popularity; Computer science; Open source software; Source code; Empirical research; World Wide Web; Code (set theory); Software; Software engineering; Join (topology); Database; Programming language; Political science","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"],"consensus_categories":[],"category_scores_codex":[0.005991622,0.0003666105,0.0003625432,0.004614401,0.0009149009,0.001638865,0.0008792484,0.001014829,0.002186837],"category_scores_gemma":[0.07276264,0.0004311647,0.0002453162,0.005796412,0.001131924,0.002802771,0.001510006,0.001579104,0.001652908],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0007542508,"about_ca_system_score_gemma":0.0006888772,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.005014406,"about_ca_topic_score_gemma":0.01101009,"domain_scores_codex":[0.9943008,0.002082036,0.0005262841,0.0008824609,0.001606352,0.0006021116],"domain_scores_gemma":[0.7745602,0.1376217,0.05000433,0.007897707,0.01957092,0.01034512],"domain_codex":null,"domain_gemma":"methods","domain_candidate":"methods","domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"observational","study_design_scores_codex":[0.0001456742,0.0002333853,0.9752545,0.00014598,0.0000516503,0.0003216402,0.008812138,0.0001877491,0.0009927535,0.0002308032,0.002001079,0.01162273],"study_design_scores_gemma":[0.000008187892,0.00009619522,0.9889431,0.00003046163,0.00001213822,0.0003433204,0.00702312,0.0007551821,0.0004282168,0.00007670617,0.002263958,0.00001936148],"study_design_candidate":"observational","study_design_consensus":"observational","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9978021,0.0001020226,0.0002386452,0.0001149828,0.000005185142,0.000021109,0.0007025828,0.00002850803,0.0009847388],"genre_scores_gemma":[0.9960168,0.0001417976,0.0007523904,0.00005483765,0.00001740175,0.0000544081,0.001664991,0.00005061478,0.001246577],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.9940084,"threshold_uncertainty_score":0.03168708,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.0645027874584888,"score_gpt":0.3501109716192732,"score_spread":0.2856081841607844,"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."}}