{"id":"W4404573117","doi":"10.5206/cjils-rcsib.v47i3.18988","title":"Lack of Collections as Data","year":2024,"lang":"en","type":"article","venue":"Canadian Journal of Information and Library Science","topic":"Web Data Mining and Analysis","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Concordia University","funders":"","keywords":"Data science; Computer science","routes":{"ca_aff":true,"ca_fund":false,"ca_venue":true,"about_ca":true,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":["scholarly_communication"],"consensus_categories":["scholarly_communication"],"category_scores_codex":[0.0004238742,0.00003346147,0.00006195895,0.0009760982,0.0001772618,0.001247844,0.001139649,0.00001285548,0.00006083733],"category_scores_gemma":[0.0001174015,0.00002675168,0.0000160035,0.002201082,0.0001596227,0.01979028,0.00009200326,0.00007195815,0.00001310145],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00001050218,"about_ca_system_score_gemma":0.003907312,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0003607814,"about_ca_topic_score_gemma":0.00008243218,"domain_scores_codex":[0.9994418,0.000009048349,0.000236505,0.00006147276,0.0001546465,0.00009649617],"domain_scores_gemma":[0.9993161,0.00002873634,0.00008476689,0.0002284415,0.00005717311,0.0002847936],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"not_applicable","study_design_scores_codex":[0.000003485657,0.000008129311,0.002573391,0.00009265853,0.00005044467,0.00005837308,0.008421472,0.0003302611,0.0001183519,0.3814232,0.3377751,0.2691451],"study_design_scores_gemma":[0.00009986203,0.00009932215,0.002231799,0.0001517394,0.00001119315,0.0004262657,0.000724937,0.1929409,0.0005033079,0.002089397,0.8006105,0.0001108285],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.1950709,0.006829061,0.4294183,0.0585466,0.006079443,0.0003066554,0.0009327395,0.0002062993,0.30261],"genre_scores_gemma":[0.9822423,0.00009246136,0.01608575,0.0008270024,0.00003809933,1.534246e-7,0.00001187964,0.000001520568,0.0007008569],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.7871714,"threshold_uncertainty_score":0.9997889,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03129558829252761,"score_gpt":0.2573358799481767,"score_spread":0.2260402916556492,"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."}}