{"id":"W2149864547","doi":"10.1109/isese.2005.1541846","title":"Cloning by accident: an empirical study of source code cloning across software systems","year":2005,"lang":"en","type":"article","venue":"","topic":"Software Engineering Research","field":"Computer Science","cited_by":79,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Waterloo","funders":"","keywords":"Cloning (programming); Computer science; Source code; Domain (mathematical analysis); Set (abstract data type); Open source; Code (set theory); Programming language; Software; Notice; Operating system; World Wide Web; Software engineering","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":[0.001272572,0.0002082507,0.0003543607,0.0001205004,0.0002038693,0.000432516,0.001722584,0.00009834388,0.00001120876],"category_scores_gemma":[0.0006983951,0.0001935772,0.00005760236,0.0006236206,0.00003612151,0.0008003699,0.0007495579,0.000339222,0.00003948084],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001263414,"about_ca_system_score_gemma":0.00005539521,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0004070424,"about_ca_topic_score_gemma":0.0001081281,"domain_scores_codex":[0.9971199,0.0001802462,0.0005083982,0.0006199414,0.0009036508,0.0006678812],"domain_scores_gemma":[0.9974494,0.001007924,0.0001037903,0.001017463,0.0001739826,0.0002474154],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.00001010125,0.0006855134,0.8851566,0.00003732324,0.00005637706,0.00001920291,0.01529282,0.08254941,0.0004175554,0.00004726908,0.003389799,0.01233801],"study_design_scores_gemma":[0.003650331,0.002302671,0.2626274,0.0001956395,0.00002269012,0.0001601731,0.00780714,0.6990212,0.007104158,0.00001629823,0.01558651,0.001505874],"study_design_candidate":"observational","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.5631083,0.0001428106,0.4358369,0.00004834052,0.0001165366,0.0001988089,0.000001640227,0.0005354085,0.00001124678],"genre_scores_gemma":[0.9804462,0.000001953536,0.01864483,0.000049438,0.0001159342,0.00003575674,0.000002736846,0.00003233007,0.0006707747],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.6225293,"threshold_uncertainty_score":0.789385,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03103695912387,"score_gpt":0.3509950186155593,"score_spread":0.3199580594916893,"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."}}