{"id":"W2749195958","doi":"10.7287/peerj.preprints.3186v1","title":"Improved query reformulation for concept location using CodeRank and document structures","year":2017,"lang":"en","type":"article","venue":"","topic":"Software Engineering Research","field":"Computer Science","cited_by":15,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Saskatchewan","funders":"","keywords":"Computer science; Query expansion; Information retrieval; Sargable; Source code; Query optimization; Web query classification; Web search query; Baseline (sea); Task (project management); Software; Code (set theory); Term (time); Query language; Quality (philosophy); Data mining; Search engine; 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":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0001540325,0.00005874301,0.00006149654,0.00003781988,0.0002546703,0.0004141131,0.0002872944,0.00003429885,0.000002387868],"category_scores_gemma":[0.0002860736,0.00004887563,0.00001220856,0.00002749782,0.0000286983,0.000647246,0.0001249798,0.00003850809,3.357811e-7],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00005431306,"about_ca_system_score_gemma":0.00003861859,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0002061743,"about_ca_topic_score_gemma":0.00001087793,"domain_scores_codex":[0.9994882,0.000006763143,0.00008560035,0.0001788564,0.0001063415,0.0001342412],"domain_scores_gemma":[0.9992809,0.0001121148,0.00004643419,0.0004110437,0.0001086914,0.00004076591],"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":[0.00007033095,0.00002916125,0.01107843,0.0003746107,0.0001076161,0.000003659679,0.002880102,0.0274484,0.05475746,0.3120432,0.0005348551,0.5906721],"study_design_scores_gemma":[0.0004859792,0.00004653204,0.04037252,0.00001250914,0.00000241554,0.000004153153,0.000007531142,0.9339162,0.01693846,0.007965296,0.0001350481,0.0001133928],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.08575505,0.00005142315,0.9135104,0.000203801,0.0001665075,0.0002222283,5.53933e-7,0.00007307692,0.00001698547],"genre_scores_gemma":[0.8318205,0.000001591702,0.1680223,0.00002085441,0.0000439674,0.000008147793,0.000001278633,0.000004273959,0.00007708927],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.9064677,"threshold_uncertainty_score":0.3993303,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02896121036955701,"score_gpt":0.3266036725533051,"score_spread":0.2976424621837481,"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."}}