{"id":"W2901220009","doi":"10.1109/icsme.2018.00086","title":"NLP2API: Query Reformulation for Code Search Using Crowdsourced Knowledge and Extra-Large Data Analytics","year":2018,"lang":"en","type":"article","venue":"","topic":"Software Engineering Research","field":"Computer Science","cited_by":9,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Saskatchewan","funders":"","keywords":"Computer science; Analytics; Upload; Code (set theory); Information retrieval; Vocabulary; Data science; Crowdsourcing; World Wide Web; Natural language; Web search query; Software; Search engine; Programming language; 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":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.00122727,0.0001119193,0.0001288201,0.0001483446,0.0002124961,0.0002170015,0.001013282,0.00007108183,0.00001379991],"category_scores_gemma":[0.0005159032,0.0001005044,0.00002114982,0.0004588782,0.00006105524,0.0006728198,0.00108735,0.0001127522,0.00001794332],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00008162672,"about_ca_system_score_gemma":0.0001077049,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00003684017,"about_ca_topic_score_gemma":0.00006914213,"domain_scores_codex":[0.9986024,0.00003433756,0.0001826911,0.0004927459,0.0002471642,0.0004406189],"domain_scores_gemma":[0.9978446,0.0005555827,0.00002365873,0.00117061,0.0002729047,0.0001326061],"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":[0.0003337564,0.001452516,0.2107303,0.002324269,0.0008390951,0.00006037407,0.01881467,0.00784591,0.07695021,0.1698425,0.04069296,0.4701134],"study_design_scores_gemma":[0.0003290956,0.00006358341,0.005878038,0.00001938416,0.000004748586,0.000008884422,0.00001526033,0.9859138,0.002000896,0.000143535,0.00548679,0.0001359701],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.1241129,0.0001131707,0.8749983,0.00008523536,0.0001281999,0.000209574,0.00001888759,0.000188812,0.0001448936],"genre_scores_gemma":[0.7971456,0.000005745708,0.2021149,0.00002641206,0.0002017105,0.000003509115,0.00001911097,0.00001806815,0.0004649534],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.9780679,"threshold_uncertainty_score":0.409845,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1780982632612534,"score_gpt":0.3950507227367746,"score_spread":0.2169524594755212,"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."}}