{"id":"W4407114865","doi":"10.22541/au.173865299.90947187/v1","title":"Clustering Textual Features for Log Summarization in Large Software Systems","year":2025,"lang":"en","type":"preprint","venue":"","topic":"Web Data Mining and Analysis","field":"Computer Science","cited_by":1,"is_retracted":false,"has_abstract":true,"ca_institutions":"Polytechnique Montréal","funders":"Natural Sciences and Engineering Research Council of Canada; Compute Canada","keywords":"Automatic summarization; Computer science; Cluster analysis; Software; Data mining; Artificial intelligence; Information retrieval; Programming language","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":[0.0004955866,0.0002013994,0.0003618586,0.0003917675,0.00008327145,0.0005250341,0.001218821,0.000258848,0.000003278206],"category_scores_gemma":[0.0001845656,0.0001857285,0.0001168021,0.0003255943,0.000008699712,0.0001691968,0.002166732,0.0002688263,0.00000407929],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00007680872,"about_ca_system_score_gemma":0.0001691432,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0006174623,"about_ca_topic_score_gemma":0.001066472,"domain_scores_codex":[0.9984415,0.00006819577,0.0003352276,0.0006917419,0.0001772848,0.0002860097],"domain_scores_gemma":[0.9987435,0.0001793619,0.0001283697,0.0008128979,0.0000948834,0.00004103792],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.00005373752,0.0005717512,0.03643323,0.00772896,0.0007869844,0.00005858401,0.003636745,0.5032233,0.00006003591,0.1551746,0.1304819,0.1617901],"study_design_scores_gemma":[0.0003029476,0.00001456162,0.001018988,0.0005708019,0.00002938423,0.000001567968,0.0001328767,0.9931563,0.00002612783,0.0004076904,0.004020475,0.0003183408],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.0001143308,0.0002882853,0.9966258,0.0002115384,0.0007154656,0.0003226027,0.0003560232,0.0002230797,0.001142874],"genre_scores_gemma":[0.4002937,0.0001318882,0.5393307,0.0007567515,0.0005939472,0.0006269641,0.003819611,0.00004015492,0.05440629],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.489933,"threshold_uncertainty_score":0.7573788,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01947680944847312,"score_gpt":0.2768511037566182,"score_spread":0.2573742943081451,"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."}}