{"id":"W2050467853","doi":"10.1145/2637002.2637025","title":"A qualitative exploration of secondary assessor relevance judging behavior","year":2014,"lang":"en","type":"article","venue":"","topic":"Handwritten Text Recognition Techniques","field":"Computer Science","cited_by":16,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Waterloo","funders":"Natural Sciences and Engineering Research Council of Canada; King Saud bin Abdulaziz University for Health Science; University of Waterloo; King Abdulaziz University; King Saud University","keywords":"Relevance (law); Think aloud protocol; Certainty; Psychology; Quality (philosophy); Test (biology); Information retrieval; Computer science; Mathematics; Epistemology; Human–computer interaction","routes":{"ca_aff":true,"ca_fund":true,"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":[0.03682908,0.000604931,0.0007456413,0.003940001,0.003442608,0.003197835,0.001314532,0.0011104,0.003550601],"category_scores_gemma":[0.1340375,0.0004811504,0.0004882467,0.002297846,0.004444747,0.001993717,0.002985039,0.001730191,0.0008216681],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.004223791,"about_ca_system_score_gemma":0.002983245,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.003027734,"about_ca_topic_score_gemma":0.003092956,"domain_scores_codex":[0.9561834,0.03204245,0.002164938,0.002626942,0.005708836,0.001273374],"domain_scores_gemma":[0.7594019,0.1943211,0.008619441,0.005026541,0.02957662,0.003054436],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"qualitative","study_design_gemma":"qualitative","study_design_scores_codex":[0.0005458119,0.0002780439,0.04205844,0.001233577,0.0000458906,0.001095119,0.8473474,0.0004330235,0.02524303,0.004638603,0.003589734,0.07349126],"study_design_scores_gemma":[0.0001129277,0.001512049,0.08339307,0.001240741,0.00008997,0.001681023,0.7766484,0.008297314,0.02938291,0.01195032,0.08526634,0.0004249889],"study_design_candidate":"qualitative","study_design_consensus":"qualitative","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9138128,0.0005014797,0.06584394,0.002054141,0.0001126692,0.001364452,0.0008058879,0.0004268264,0.0150778],"genre_scores_gemma":[0.9695458,0.0002703734,0.02287753,0.0004651866,0.00003850449,0.001313587,0.0002568081,0.0001215183,0.005110711],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.03682908,"threshold_uncertainty_score":0.1947731,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.05938001020438095,"score_gpt":0.3608155720620532,"score_spread":0.3014355618576723,"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."}}