{"id":"W2170783728","doi":"10.1002/meet.14505001072","title":"Exact versus estimated pruning of subject hierarchies","year":2013,"lang":"en","type":"article","venue":"Proceedings of the American Society for Information Science and Technology","topic":"Advanced Text Analysis Techniques","field":"Computer Science","cited_by":1,"is_retracted":false,"has_abstract":true,"ca_institutions":"McGill University","funders":"","keywords":"Subject (documents); Hierarchy; Computer science; Pruning; Data science; Information retrieval; Visualization; Artificial intelligence; Theoretical computer science; World Wide Web","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":[0.0005056473,0.00008607887,0.0001952992,0.0003383522,0.0002340968,0.0000851187,0.001284532,0.00003345607,3.904374e-7],"category_scores_gemma":[0.000589954,0.00006195955,0.00008458207,0.004444062,0.002632728,0.003789337,0.0005188706,0.00009098645,6.841588e-7],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00004606785,"about_ca_system_score_gemma":0.00008916152,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.000025952,"about_ca_topic_score_gemma":1.914906e-7,"domain_scores_codex":[0.9990062,0.000001039333,0.000282494,0.0001430642,0.0003308336,0.0002363686],"domain_scores_gemma":[0.9976203,0.00006125419,0.0007339889,0.0001780488,0.001376865,0.00002947764],"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":[0.00001637144,0.00002727428,0.006854758,0.0001081618,0.00006093709,5.84179e-9,0.003056977,0.00001154653,0.2233659,0.3139222,0.001224189,0.4513516],"study_design_scores_gemma":[0.0007099812,0.0008270184,0.008665512,0.00008075269,0.0000387709,0.00001050661,0.009489328,0.1068347,0.8209304,0.04972459,0.002323577,0.0003648543],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9642922,0.00001931768,0.03215159,0.002095501,0.00002866089,0.0003966696,0.000001705333,0.000245211,0.0007691302],"genre_scores_gemma":[0.8032343,0.00003985234,0.1965867,0.0000853435,0.000002034662,0.00004574939,1.837854e-7,0.000001928063,0.000003904264],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.5975644,"threshold_uncertainty_score":0.9700395,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01388614426442249,"score_gpt":0.2833045487715939,"score_spread":0.2694184045071714,"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."}}