{"id":"W4285397054","doi":"10.48550/arxiv.2207.05132","title":"Dev2vec: Representing Domain Expertise of Developers in an Embedding Space","year":2022,"lang":"en","type":"preprint","venue":"arXiv (Cornell University)","topic":"Software Engineering Research","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Polytechnique Montréal","funders":"","keywords":"Embedding; Computer science; Domain (mathematical analysis); Space (punctuation); Software; Data science; Exploit; Software engineering; World Wide Web; Information retrieval; Artificial intelligence; Computer security; Programming language; 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":[],"consensus_categories":[],"category_scores_codex":[0.001096674,0.002492619,0.0007542393,0.00296574,0.0004142224,0.00122155,0.0008238637,0.001111314,0.002206369],"category_scores_gemma":[0.004681996,0.0003478313,0.0007580265,0.002355058,0.0006926034,0.001818323,0.001188617,0.001768685,0.001556446],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0006956712,"about_ca_system_score_gemma":0.0008603517,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.005718752,"about_ca_topic_score_gemma":0.008693486,"domain_scores_codex":[0.9987832,0.0003753433,0.00009770634,0.0003663925,0.0002460301,0.0001313299],"domain_scores_gemma":[0.9974336,0.001358164,0.0002027705,0.0003109399,0.0005787634,0.0001156954],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0008141531,0.0006678882,0.02479515,0.001020308,0.0003493796,0.0005593431,0.0009761519,0.09570122,0.02055867,0.006957444,0.08435275,0.7632475],"study_design_scores_gemma":[0.00005010658,0.0002299418,0.006674001,0.00009042429,0.00006923722,0.0003038082,0.0002672913,0.9618571,0.008813921,0.008418049,0.01316423,0.00006189834],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.3006064,0.003819528,0.6607438,0.001533976,0.001098461,0.0003427685,0.0169347,0.009409578,0.005510857],"genre_scores_gemma":[0.792918,0.001450765,0.1606888,0.0004462922,0.0004187713,0.000467233,0.03627794,0.000590231,0.006742003],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.005718752,"threshold_uncertainty_score":0.0113709,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.08073216102822943,"score_gpt":0.2387423714395805,"score_spread":0.158010210411351,"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."}}