{"id":"W1512615396","doi":"10.1002/smr.1662","title":"Big data clone detection using classical detectors: an exploratory study","year":2014,"lang":"en","type":"article","venue":"Journal of Software Evolution and Process","topic":"Software Engineering Research","field":"Computer Science","cited_by":20,"is_retracted":false,"has_abstract":true,"ca_institutions":"Queen's University; University of Saskatchewan","funders":"","keywords":"Computer science; Scalability; Big data; Source code; Code (set theory); clone (Java method); Data mining; Data science; Database; 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":[0.01469192,0.001065088,0.0007335313,0.005555054,0.0009625353,0.002107122,0.002528153,0.001294691,0.0004681821],"category_scores_gemma":[0.06631155,0.0005508933,0.001287578,0.003306074,0.001770532,0.003194642,0.002114549,0.001918138,0.0003090126],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0010199,"about_ca_system_score_gemma":0.0008647224,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002357755,"about_ca_topic_score_gemma":0.002668257,"domain_scores_codex":[0.985274,0.007457813,0.000746719,0.001844381,0.004210997,0.0004661053],"domain_scores_gemma":[0.8066698,0.1538436,0.008957678,0.01487631,0.01419676,0.00145587],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"observational","study_design_scores_codex":[0.001910929,0.005127853,0.6314976,0.002352065,0.001106537,0.002265377,0.0102344,0.03833549,0.02124615,0.008557207,0.009568627,0.2677978],"study_design_scores_gemma":[0.0003265861,0.004704618,0.2176282,0.0006053961,0.0006314742,0.00439697,0.01198088,0.6684609,0.05301902,0.01885118,0.01905975,0.0003350844],"study_design_candidate":"observational","study_design_consensus":"observational","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9474265,0.0007315926,0.0475673,0.0003483577,0.00003201449,0.0005837937,0.0009030233,0.0007547071,0.001652815],"genre_scores_gemma":[0.9297568,0.000247884,0.06767689,0.0001216364,0.00003417661,0.0002762415,0.001361359,0.000128549,0.0003966207],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.01469192,"threshold_uncertainty_score":0.07769924,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.07457598398553884,"score_gpt":0.3182856481568901,"score_spread":0.2437096641713513,"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."}}