{"id":"W3162439607","doi":"10.31234/osf.io/9a52q","title":"Content matters: Measures of contextual diversity must consider semantic content","year":2021,"lang":"en","type":"preprint","venue":"","topic":"Topic Modeling","field":"Computer Science","cited_by":8,"is_retracted":false,"has_abstract":true,"ca_institutions":"McGill University","funders":"","keywords":"Optimal distinctiveness theory; Diversity (politics); Word (group theory); Computer science; Natural language processing; Word lists by frequency; Context (archaeology); Artificial intelligence; Variance (accounting); Linguistics; Content (measure theory); Psychology; Mathematics; Social psychology; Sociology; History","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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.006939542,0.001054719,0.001440208,0.004365214,0.001447191,0.004795195,0.001108462,0.001906744,0.003212787],"category_scores_gemma":[0.06105398,0.000610063,0.001162906,0.004798966,0.003238119,0.0157422,0.004494881,0.002755318,0.0008332744],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001264186,"about_ca_system_score_gemma":0.00102927,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002438529,"about_ca_topic_score_gemma":0.001754628,"domain_scores_codex":[0.9945746,0.002028448,0.000494391,0.001509346,0.001068281,0.0003249597],"domain_scores_gemma":[0.9594615,0.02697267,0.002917256,0.006807539,0.002616669,0.001224366],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.001407515,0.0006342411,0.2478226,0.001685839,0.001771091,0.0004064535,0.006857827,0.04520267,0.02527175,0.2332044,0.008341383,0.4273942],"study_design_scores_gemma":[0.000108316,0.0005640865,0.1349986,0.0003160116,0.0004662498,0.0007749861,0.002833919,0.09844925,0.008978515,0.7341842,0.0180395,0.000286351],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.4053695,0.003865717,0.5535704,0.002764612,0.0003670576,0.0005165251,0.002209147,0.0008459236,0.03049101],"genre_scores_gemma":[0.9331756,0.0004719337,0.06317244,0.0004477687,0.0003223927,0.0003113876,0.001122523,0.000221941,0.0007539984],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.006939542,"threshold_uncertainty_score":0.03670025,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.2554150969675804,"score_gpt":0.2713124592966478,"score_spread":0.01589736232906747,"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."}}