{"id":"W4298195797","doi":"","title":"1 Exact versus Estimated Pruning of Subject Hierarchies","year":2016,"lang":"en","type":"article","venue":"","topic":"Advanced Text Analysis Techniques","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"McGill University","funders":"","keywords":"Pruning; Subject (documents); Computer science; Artificial intelligence; Horticulture; World Wide Web; Biology","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":[0.0001364849,0.00008654129,0.0001459655,0.000162611,0.0000352495,0.00001897884,0.0005779197,0.00002652348,0.00004028482],"category_scores_gemma":[0.0001295958,0.00005275307,0.00005566634,0.0003890567,0.00007169991,0.0005287911,0.0002010677,0.00003440342,0.0000190841],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00002742637,"about_ca_system_score_gemma":0.00002719178,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00002063437,"about_ca_topic_score_gemma":0.00001089223,"domain_scores_codex":[0.9992487,0.00003109609,0.00017108,0.0002137063,0.000168566,0.0001668219],"domain_scores_gemma":[0.9990298,0.0002809223,0.00008419311,0.000499305,0.00006725834,0.00003850507],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"bench_or_experimental","study_design_scores_codex":[0.00004851832,0.00005958337,0.003240754,0.00001094822,0.00007936242,0.00001062994,0.0002356721,0.00003545784,0.1862096,0.2942726,0.0006486426,0.5151482],"study_design_scores_gemma":[0.0008574249,0.0003033562,0.00454157,0.00009415378,0.00001539951,0.000004220375,0.0000139052,0.01130217,0.953509,0.02783447,0.001204773,0.0003194973],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.03032125,0.0000318415,0.9607035,0.0002430128,0.00004354958,0.00004889771,4.632737e-7,0.0005475326,0.008059986],"genre_scores_gemma":[0.6785703,0.00001778319,0.3210546,0.00001061991,0.000005819202,0.000003473137,2.120175e-7,0.000004015332,0.0003331311],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.7672994,"threshold_uncertainty_score":0.2151208,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03281499216325376,"score_gpt":0.3134140566580464,"score_spread":0.2805990644947927,"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."}}