{"id":"W3176244545","doi":"10.48550/arxiv.2001.01424","title":"Cross-Dataset Design Discussion Mining","year":2020,"lang":"en","type":"preprint","venue":"Figshare","topic":"Data Mining Algorithms and Applications","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Victoria","funders":"","keywords":"Computer science; Artifact (error); Code refactoring; Relevance (law); Context (archaeology); Documentation; Task (project management); Software; Data mining; Machine learning; Software engineering; Artificial intelligence; Data science; Engineering; Systems engineering; 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":[0.009221399,0.001284509,0.000813826,0.01175557,0.002281151,0.003508226,0.003203991,0.002357059,0.01123483],"category_scores_gemma":[0.03926051,0.000491399,0.001759789,0.008812296,0.0008070778,0.002942323,0.003908603,0.002194398,0.006943631],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001766081,"about_ca_system_score_gemma":0.002534589,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.003360288,"about_ca_topic_score_gemma":0.007000553,"domain_scores_codex":[0.9864738,0.003907565,0.001733937,0.003473253,0.003733171,0.0006782398],"domain_scores_gemma":[0.960457,0.01606946,0.003260513,0.01170697,0.007439977,0.00106616],"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":[0.001293159,0.002277848,0.1080469,0.006455879,0.00104853,0.0009755153,0.003266382,0.007496779,0.01195364,0.01706084,0.3179864,0.5221382],"study_design_scores_gemma":[0.0004088651,0.0005280724,0.09427492,0.0009008353,0.0004966996,0.001216501,0.003716716,0.03235199,0.02660255,0.01729247,0.8220081,0.0002023062],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"dataset","genre_gemma":"empirical","genre_scores_codex":[0.2545541,0.006150856,0.1214095,0.003006983,0.001001206,0.006306112,0.5306028,0.01155151,0.06541696],"genre_scores_gemma":[0.231701,0.0007351093,0.147466,0.001061635,0.0002097154,0.006706296,0.5935538,0.0009463055,0.01762016],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.01175557,"threshold_uncertainty_score":0.04876804,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1440692337339377,"score_gpt":0.346395833574453,"score_spread":0.2023265998405153,"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."}}