{"id":"W2100848433","doi":"10.1002/meet.1450400107","title":"Haystacks and hypotheses","year":2003,"lang":"en","type":"article","venue":"Proceedings of the American Society for Information Science and Technology","topic":"Biomedical Text Mining and Ontologies","field":"Biochemistry, Genetics and Molecular Biology","cited_by":2,"is_retracted":false,"has_abstract":true,"ca_institutions":"National Research Council Canada","funders":"","keywords":"Haystack; Process (computing); Subject (documents); Computer science; Information retrieval; Term (time); Data science; Scientific literature; Artificial intelligence; Library science; 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":["sts"],"consensus_categories":[],"category_scores_codex":[0.0004064992,0.00005053047,0.00008059602,0.00004782175,0.0001705657,0.00002594211,0.0001803862,0.00004872753,1.740016e-7],"category_scores_gemma":[0.001116662,0.00003353628,0.00003201502,0.0005186946,0.003265177,0.00002481791,0.0001036516,0.00003998825,1.324254e-7],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.000005995895,"about_ca_system_score_gemma":0.00005878759,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.000001803497,"about_ca_topic_score_gemma":1.464748e-7,"domain_scores_codex":[0.9995517,7.113088e-7,0.0001032199,0.00009568139,0.0001061898,0.0001425275],"domain_scores_gemma":[0.9994259,0.000009463183,0.0001582843,0.00005664901,0.000324657,0.00002498443],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"bench_or_experimental","study_design_gemma":"bench_or_experimental","study_design_scores_codex":[0.00002672945,0.00002246806,0.03197075,0.0001023742,0.00004585912,6.138651e-9,0.0007831558,2.62801e-7,0.6463667,0.04908052,0.007586292,0.2640148],"study_design_scores_gemma":[0.0004535613,0.0006302774,0.004867062,0.00001737928,0.00001950044,0.00003886785,0.0197448,0.000113157,0.5100621,0.004411045,0.4594588,0.0001834849],"study_design_candidate":"bench_or_experimental","study_design_consensus":"bench_or_experimental","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9976292,0.00009530994,0.0001922359,0.001100777,0.00001960194,0.00009912288,0.000003700332,0.00001473609,0.0008453153],"genre_scores_gemma":[0.9900551,0.0001638733,0.009247238,0.0004852808,0.000005013298,0.00001544644,2.880929e-7,0.000001709461,0.00002604096],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.4518725,"threshold_uncertainty_score":0.9994473,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01011866201815301,"score_gpt":0.2538281088009441,"score_spread":0.2437094467827911,"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."}}