{"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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.02483297,0.00177624,0.002424949,0.007620557,0.0007223299,0.004341996,0.004859469,0.003730822,0.004679599],"category_scores_gemma":[0.1572492,0.001461755,0.001525578,0.004513463,0.001310072,0.01155841,0.005304649,0.002820756,0.001905094],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0006134231,"about_ca_system_score_gemma":0.002264888,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0005840245,"about_ca_topic_score_gemma":0.0008355726,"domain_scores_codex":[0.96345,0.0137332,0.005404716,0.003192573,0.01315778,0.001061727],"domain_scores_gemma":[0.7228567,0.1801267,0.01349512,0.05740023,0.02430235,0.00181892],"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":[0.001336317,0.001347383,0.02575947,0.001756761,0.0002533516,0.003929635,0.006258553,0.01607797,0.134793,0.01424227,0.01245481,0.7817905],"study_design_scores_gemma":[0.001265533,0.003022842,0.03506932,0.002004911,0.0008417094,0.01129515,0.00175504,0.5775071,0.2315141,0.0381539,0.09661034,0.0009600685],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.08979205,0.0004130217,0.8490395,0.0009718825,0.00009831142,0.0006335969,0.000551998,0.05547965,0.003020024],"genre_scores_gemma":[0.1650843,0.0001723036,0.829683,0.0001853249,0.00006527676,0.0004253473,0.001114509,0.002017782,0.001252165],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.02483297,"threshold_uncertainty_score":0.1313309,"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."}}