{"id":"W4281491548","doi":"10.1098/rsos.211189","title":"Human-supervised clustering of multidimensional data using crowdsourcing","year":2022,"lang":"en","type":"article","venue":"Royal Society Open Science","topic":"Mobile Crowdsensing and Crowdsourcing","field":"Computer Science","cited_by":3,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Manitoba; McGill University","funders":"Canadian Institutes of Health Research; Genome Canada","keywords":"Cluster analysis; Crowdsourcing; Computer science; Data mining; Metric (unit); Task (project management); Machine learning; Artificial intelligence; Clustering high-dimensional data","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":"codex-gemma-dda1882f352a","candidate_categories":["sts","open_science"],"consensus_categories":["open_science"],"category_scores_codex":[0.004794118,0.0001857942,0.0002968547,0.00007610548,0.003702267,0.0005928286,0.01038563,0.00003452354,0.00008721399],"category_scores_gemma":[0.00008107138,0.0001955946,0.0001079069,0.001368566,0.0004658601,0.001256852,0.0315602,0.0003267934,0.000003173438],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0002392291,"about_ca_system_score_gemma":0.000615843,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00142049,"about_ca_topic_score_gemma":0.00001113033,"domain_scores_codex":[0.9963769,0.0001269635,0.0004354831,0.001180313,0.001284824,0.0005954752],"domain_scores_gemma":[0.9971247,0.00009022019,0.000253912,0.0022011,0.0001637763,0.0001662957],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.000009608792,0.000276054,0.002169899,0.00005324332,0.00004083054,0.00002117563,0.00834921,0.6370818,0.3356202,0.003229989,0.002713459,0.01043448],"study_design_scores_gemma":[0.0004192749,0.00004845851,0.0009498915,0.00003510448,0.000008057741,0.00003032588,0.001137782,0.9938988,0.002430289,0.00006338589,0.0007255388,0.0002531117],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.6784208,0.000104262,0.3186376,0.0002330139,0.0005976675,0.000435157,0.00002611899,0.0001130504,0.001432359],"genre_scores_gemma":[0.7863156,5.415819e-7,0.2130907,0.0003159695,0.00003536122,0.00000658034,0.000005248158,0.0000127461,0.000217203],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.3568169,"threshold_uncertainty_score":0.9975948,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.06761932082304395,"score_gpt":0.3258640248512842,"score_spread":0.2582447040282402,"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."}}