{"id":"W2155056246","doi":"10.1093/comjnl/bxs018","title":"A Source Code Similarity System for Plagiarism Detection","year":2012,"lang":"en","type":"article","venue":"The Computer Journal","topic":"Academic integrity and plagiarism","field":"Social Sciences","cited_by":109,"is_retracted":false,"has_abstract":true,"ca_institutions":"Athabasca University","funders":"","keywords":"Library science; Computer science; Similarity (geometry); Information retrieval; Source code; World Wide Web; Database; Operating system; Artificial intelligence","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":["research_integrity"],"consensus_categories":[],"category_scores_codex":[0.005108945,0.001036508,0.001207807,0.009843413,0.001228029,0.001669763,0.001529963,0.00176047,0.00640064],"category_scores_gemma":[0.02350937,0.0006553412,0.0007502857,0.003639013,0.0005802275,0.003279973,0.002210915,0.001460611,0.005520111],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0007553934,"about_ca_system_score_gemma":0.001960662,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001870944,"about_ca_topic_score_gemma":0.001491275,"domain_scores_codex":[0.9948626,0.0007821222,0.0009842059,0.001118965,0.002089581,0.0001625079],"domain_scores_gemma":[0.9809137,0.004467306,0.002479027,0.003502612,0.007959499,0.0006779253],"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":[0.0007662558,0.0006921408,0.01787493,0.001009419,0.0001821269,0.0006403974,0.001175553,0.002919947,0.08184329,0.004546536,0.03031723,0.8580322],"study_design_scores_gemma":[0.0006713851,0.002359948,0.1062169,0.0004080355,0.0004427875,0.006117919,0.0008597667,0.4040574,0.3362769,0.01063942,0.131169,0.0007805231],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.04420534,0.0003884406,0.7694525,0.0003423735,0.0001423209,0.002178854,0.002676177,0.176255,0.004358974],"genre_scores_gemma":[0.1919053,0.0002676209,0.7892629,0.0001984278,0.0001163937,0.001286733,0.007908975,0.002571124,0.006482523],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.9982395,"threshold_uncertainty_score":0.02701896,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03357069603471791,"score_gpt":0.294971855487683,"score_spread":0.2614011594529651,"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."}}