{"id":"W2142459460","doi":"","title":"Evaluating Roget's Thesauri","year":2008,"lang":"en","type":"article","venue":"","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":25,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Ottawa","funders":"","keywords":"Thesaurus; WordNet; Computer science; Sentence; Information retrieval; Natural language processing; Artificial intelligence; Synonym (taxonomy); 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":[0.0002119019,0.00006523856,0.00006450713,0.00004580251,0.0001208516,0.00004806552,0.0006848737,0.00003130632,0.00003095671],"category_scores_gemma":[0.00008546217,0.00004783573,0.00002744407,0.0002120158,0.00002432185,0.0003961438,0.000196261,0.00008528362,0.00005182483],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00001681883,"about_ca_system_score_gemma":0.00004665429,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00001883406,"about_ca_topic_score_gemma":9.689074e-7,"domain_scores_codex":[0.9993076,0.00002774417,0.00009480114,0.0001946447,0.0002288132,0.0001463349],"domain_scores_gemma":[0.9994875,0.00004448774,0.00003371565,0.0003370749,0.00006272101,0.00003455973],"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":[0.000003476112,0.000067074,0.001428786,0.00001573491,0.0000122619,0.0001412662,0.002067604,0.00001156008,0.02765903,0.4657761,0.006780482,0.4960366],"study_design_scores_gemma":[0.0004016061,0.0002585095,0.0007839666,0.00004890515,0.000005325387,0.0009981963,0.00002261856,0.1469072,0.5491476,0.2979831,0.002747635,0.0006954008],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.0213383,0.00115552,0.9650236,0.0008196936,0.00007822372,0.00008270905,9.672408e-8,0.001782439,0.009719362],"genre_scores_gemma":[0.3905881,0.000003568878,0.6083072,0.0003930153,0.00002064217,0.000003976795,1.610533e-7,0.000002756359,0.0006805534],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.5214885,"threshold_uncertainty_score":0.1950684,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.06413836483233207,"score_gpt":0.3523428756106328,"score_spread":0.2882045107783007,"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."}}