{"id":"W2107321288","doi":"10.1109/icsm.2005.52","title":"Improved tool support for the investigation of duplication in software","year":2005,"lang":"en","type":"article","venue":"","topic":"Software Engineering Research","field":"Computer Science","cited_by":64,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Waterloo","funders":"","keywords":"Computer science; Software system; Software; Software engineering; Code (set theory); Software maintenance; Program comprehension; Gene duplication; Set (abstract data type); Software construction; Cloning (programming); Software evolution; Programming language","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":[0.0004199192,0.00004166659,0.00004981576,0.00006375637,0.00002068848,0.00002485746,0.0004276788,0.00002667075,0.000008722565],"category_scores_gemma":[0.000592507,0.00003035634,0.00002071124,0.0002545686,0.00001902119,0.000232035,0.00006466834,0.00004964057,0.000005983793],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0000372497,"about_ca_system_score_gemma":0.00007078343,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00002337597,"about_ca_topic_score_gemma":0.00001736819,"domain_scores_codex":[0.9994802,0.000007644584,0.0001433549,0.0001304749,0.0001180125,0.0001203509],"domain_scores_gemma":[0.9987532,0.0007779392,0.00002715359,0.0003431213,0.0000788982,0.00001963571],"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":[0.000020026,0.00008355213,0.1254918,0.000148816,0.00002696502,4.709613e-7,0.002318873,0.01216703,0.03748306,0.03828417,0.006232006,0.7777433],"study_design_scores_gemma":[0.0003872043,0.00007381921,0.2469795,0.000007692973,0.000001679989,0.00000181156,0.000008361838,0.69446,0.05481852,0.0009474433,0.002211755,0.0001021928],"study_design_candidate":"design_other","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.05923572,0.00001722555,0.9387443,0.001526646,0.00003805925,0.0003307317,7.99653e-7,0.0000943827,0.00001214137],"genre_scores_gemma":[0.7187471,0.00000243096,0.2806863,0.0001149498,0.00002275265,0.0001229076,0.000002267379,0.000004448739,0.0002968299],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.7776411,"threshold_uncertainty_score":0.1237895,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02047931845975989,"score_gpt":0.2692938742884374,"score_spread":0.2488145558286775,"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."}}