{"id":"W2602793737","doi":"10.1109/msr.2017.50","title":"Rediscovery Datasets: Connecting Duplicate Reports","year":2017,"lang":"en","type":"article","venue":"","topic":"Software Engineering Research","field":"Computer Science","cited_by":9,"is_retracted":false,"has_abstract":true,"ca_institutions":"Toronto Metropolitan University","funders":"","keywords":"Software bug; Open source software; Software; Open source; Tracking (education); Tracking system","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":["metaresearch"],"consensus_categories":[],"category_scores_codex":[0.004448,0.001822569,0.001210931,0.01289206,0.001696954,0.002561294,0.004312382,0.003074852,0.001796412],"category_scores_gemma":[0.02722301,0.0007577231,0.001530587,0.009278233,0.0007684559,0.002755386,0.004170724,0.001944169,0.002883555],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001140316,"about_ca_system_score_gemma":0.002472981,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01465416,"about_ca_topic_score_gemma":0.01594439,"domain_scores_codex":[0.9934055,0.0008372247,0.001346467,0.001702033,0.002229437,0.0004793198],"domain_scores_gemma":[0.9780815,0.00580333,0.003883149,0.006155618,0.00524856,0.0008278618],"domain_codex":null,"domain_gemma":"reproducibility","domain_candidate":"reproducibility","domain_consensus":null,"study_design_codex":"not_applicable","study_design_gemma":"not_applicable","study_design_scores_codex":[0.00110633,0.001135658,0.1433474,0.003934867,0.001072653,0.002013872,0.001977317,0.01597343,0.006495137,0.00856181,0.6590486,0.1553329],"study_design_scores_gemma":[0.0005355195,0.0004015341,0.1383187,0.0008815274,0.0004961638,0.003228078,0.002546943,0.08962923,0.02093469,0.01536166,0.7272929,0.0003731811],"study_design_candidate":"not_applicable","study_design_consensus":"not_applicable","genre_codex":"dataset","genre_gemma":"methods","genre_scores_codex":[0.1021823,0.003570465,0.01845075,0.00156028,0.0005024323,0.0007550471,0.8513795,0.01619831,0.005400918],"genre_scores_gemma":[0.05151656,0.0005538797,0.02766798,0.0002240368,0.00008908944,0.0005620255,0.9175683,0.000436029,0.001382175],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.995552,"threshold_uncertainty_score":0.02913773,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03030703116912421,"score_gpt":0.3088519965327439,"score_spread":0.2785449653636197,"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."}}