{"id":"W2158324307","doi":"10.1109/icsm.2007.4362614","title":"Mining the Lexicon Used by Programmers during Sofware Evolution","year":2007,"lang":"en","type":"article","venue":"","topic":"Software Engineering Research","field":"Computer Science","cited_by":57,"is_retracted":false,"has_abstract":true,"ca_institutions":"Université de Montréal; Polytechnique Montréal","funders":"Natural Sciences and Engineering Research Council of Canada; Canada Research Chairs","keywords":"Identifier; Lexicon; Computer science; Program comprehension; Documentation; Process (computing); Task (project management); Comprehension; Natural language processing; Artificial intelligence; Information retrieval; Programming language; Software; Software system","routes":{"ca_aff":true,"ca_fund":true,"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":[0.000645043,0.00008568355,0.00005999913,0.00007954094,0.0001480753,0.0001358698,0.0006696894,0.00004242401,0.00001013732],"category_scores_gemma":[0.0002047799,0.00006183748,0.00003389013,0.0004974591,0.0000364109,0.0002476261,0.0001772079,0.0001402747,0.00003122879],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001264108,"about_ca_system_score_gemma":0.00003966201,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0000746044,"about_ca_topic_score_gemma":0.00001732636,"domain_scores_codex":[0.9987912,0.00001944424,0.0001232324,0.0002349442,0.0003750715,0.0004561656],"domain_scores_gemma":[0.9990508,0.0004060804,0.00002183346,0.0003954318,0.00004052866,0.00008535373],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"observational","study_design_scores_codex":[0.00004574098,0.000307742,0.580292,0.0001756771,0.0001570878,0.0001391686,0.01214901,0.001897706,0.09812521,0.02800508,0.01518381,0.2635217],"study_design_scores_gemma":[0.002262463,0.0003816284,0.6751513,0.0001273952,0.000012876,0.0001477139,0.002600328,0.08049414,0.2237878,0.0006423297,0.01297556,0.001416389],"study_design_candidate":"observational","study_design_consensus":"observational","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.3603964,0.00007055827,0.6382179,0.0004005317,0.0001260232,0.0001189439,1.444396e-7,0.0004395247,0.0002299217],"genre_scores_gemma":[0.9819727,5.661938e-7,0.01718364,0.00002288351,0.00004368648,0.00001129528,7.970008e-7,0.00000941853,0.0007550335],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.6215762,"threshold_uncertainty_score":0.2521659,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01567019451879826,"score_gpt":0.2642007931308423,"score_spread":0.248530598612044,"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."}}