{"id":"W2545762469","doi":"10.1002/smr.1821","title":"Detecting duplicate bug reports with software engineering domain knowledge","year":2016,"lang":"en","type":"article","venue":"Journal of Software Evolution and Process","topic":"Software Engineering Research","field":"Computer Science","cited_by":61,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Alberta","funders":"","keywords":"Computer science; Domain (mathematical analysis); Software engineering; Domain knowledge; Context (archaeology); Data science; Software; Domain engineering; Software mining; Information retrieval; Data mining; Software development; Software construction","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":[0.000927548,0.0002007302,0.0002709815,0.0003669881,0.0001344578,0.0001121718,0.0004280857,0.00009263391,0.00000611713],"category_scores_gemma":[0.002740011,0.0001294341,0.000063479,0.0006248805,0.00005450508,0.0009834459,0.0001272692,0.0002955118,0.000005292874],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001877462,"about_ca_system_score_gemma":0.0002907138,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.000002358571,"about_ca_topic_score_gemma":0.000001774382,"domain_scores_codex":[0.998312,0.00003195142,0.0004647479,0.0003174636,0.0004868982,0.0003869471],"domain_scores_gemma":[0.9975784,0.0008044543,0.0003348669,0.0003141188,0.0006959687,0.0002722257],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"observational","study_design_scores_codex":[0.000202209,0.0002618191,0.8301801,0.001269937,0.0003084308,0.001274923,0.003190134,0.002834085,0.004350347,0.0008699266,0.00100718,0.1542509],"study_design_scores_gemma":[0.007922566,0.003480504,0.8624274,0.009151154,0.0001418688,0.05065598,0.0003267662,0.003541597,0.01911026,0.01666558,0.02305322,0.003523126],"study_design_candidate":"observational","study_design_consensus":"observational","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.2516186,0.001219861,0.7463815,0.0001723342,0.0002913404,0.00008060592,7.652931e-7,0.0002311992,0.000003745283],"genre_scores_gemma":[0.9066812,0.0000257389,0.0929713,0.00001175422,0.000205483,0.00001062277,1.41757e-7,0.00002675616,0.00006701296],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.6550626,"threshold_uncertainty_score":0.527817,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.008544446977499068,"score_gpt":0.2413264381142093,"score_spread":0.2327819911367102,"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."}}