{"id":"W4409441986","doi":"10.1007/s10664-025-10655-9","title":"Predicting long time contributors with knowledge units of programming languages: an empirical study","year":2025,"lang":"en","type":"article","venue":"Empirical Software Engineering","topic":"Software Engineering Research","field":"Computer Science","cited_by":3,"is_retracted":false,"has_abstract":false,"ca_institutions":"Queen's University","funders":"","keywords":"Computer science; Empirical research; Programming language; Natural language processing; Data science; Artificial intelligence; Statistics; Mathematics","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","bibliometrics"],"consensus_categories":[],"category_scores_codex":[0.003626236,0.0003620716,0.0003419942,0.002126383,0.001086786,0.001794665,0.001378039,0.001554997,0.006747188],"category_scores_gemma":[0.05783837,0.0005234166,0.0004095845,0.001481421,0.0006475766,0.003115134,0.0016617,0.002251112,0.001635952],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0005439252,"about_ca_system_score_gemma":0.001004037,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.008350407,"about_ca_topic_score_gemma":0.01048862,"domain_scores_codex":[0.9978495,0.0009449731,0.0001466306,0.0003735636,0.0004637495,0.0002215779],"domain_scores_gemma":[0.8323547,0.1171499,0.02288959,0.007413031,0.006916657,0.01327622],"domain_codex":null,"domain_gemma":"incentives","domain_candidate":"incentives","domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"observational","study_design_scores_codex":[0.0002395091,0.001162088,0.9873892,0.00001722439,0.00004265715,0.000152806,0.002072074,0.0002926335,0.0002712743,0.0002207273,0.0003547786,0.007784995],"study_design_scores_gemma":[0.00004918973,0.0005864514,0.9766533,0.00004047132,0.0001096838,0.0004405193,0.006398903,0.01229445,0.0008750968,0.0007731084,0.001740977,0.00003795618],"study_design_candidate":"observational","study_design_consensus":"observational","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9991847,0.00004571381,0.0003041425,0.00006830019,0.000004647723,0.000009010585,0.00008072743,0.000007384114,0.0002954191],"genre_scores_gemma":[0.9975005,0.00007342078,0.0004092199,0.00002688087,0.00001508872,0.00001863666,0.0002543497,0.00001473158,0.001687211],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.9978736,"threshold_uncertainty_score":0.02257156,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01772367776384217,"score_gpt":0.3199314896089665,"score_spread":0.3022078118451244,"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."}}