{"id":"W1979991500","doi":"10.1007/s10664-012-9224-x","title":"Configuring latent Dirichlet allocation based feature location","year":2012,"lang":"en","type":"article","venue":"Empirical Software Engineering","topic":"Software Engineering Research","field":"Computer Science","cited_by":92,"is_retracted":false,"has_abstract":false,"ca_institutions":"University of Waterloo","funders":"U.S. Department of Education; National Science Foundation","keywords":"Latent Dirichlet allocation; Computer science; Feature (linguistics); Source code; Heuristics; Context (archaeology); Java; Artificial intelligence; Topic model; Code (set theory); Measure (data warehouse); Data mining; Natural language processing; Information retrieval; 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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.004691952,0.001172293,0.001994057,0.001771715,0.001198845,0.002288966,0.003565405,0.002996026,0.005946101],"category_scores_gemma":[0.02617802,0.001117412,0.001562748,0.002036664,0.001130444,0.004341395,0.004427487,0.0025514,0.00408932],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001247305,"about_ca_system_score_gemma":0.001734929,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.005126292,"about_ca_topic_score_gemma":0.006945486,"domain_scores_codex":[0.9952041,0.002340378,0.0002778222,0.001211521,0.0005941085,0.0003720744],"domain_scores_gemma":[0.9893944,0.006382376,0.0002859529,0.002386235,0.001230207,0.0003208873],"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":[0.002806083,0.0006807047,0.01138373,0.0002170039,0.000360265,0.0003402931,0.0006890604,0.1710555,0.0198756,0.01983182,0.01693756,0.7558224],"study_design_scores_gemma":[0.00009839551,0.00005694141,0.0006162566,0.00001183574,0.00005144429,0.00008170229,0.00009459545,0.9660085,0.006740224,0.02454561,0.001665583,0.00002891898],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.02798516,0.0001884386,0.9646391,0.000317795,0.0001225927,0.00008263373,0.0003234452,0.00535415,0.0009866538],"genre_scores_gemma":[0.4894229,0.0001269988,0.503398,0.0003601254,0.0001442928,0.0004010892,0.002122656,0.0009832985,0.003040629],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.005946101,"threshold_uncertainty_score":0.02481377,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.0249021583454907,"score_gpt":0.2743756658225636,"score_spread":0.2494735074770729,"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."}}