{"id":"W4395112800","doi":"10.1038/s41597-024-03275-3","title":"A new commercial boundary dataset for metropolitan areas in the USA and Canada, built from open data","year":2024,"lang":"en","type":"article","venue":"Scientific Data","topic":"Human Mobility and Location-Based Analysis","field":"Social Sciences","cited_by":1,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Toronto","funders":"University of Toronto","keywords":"Metropolitan area; Boundary (topology); Geography; Regional science; Cluster analysis; Data collection; Data science; Segmentation; Economic geography; Cartography; Computer science; Statistics","routes":{"ca_aff":true,"ca_fund":true,"ca_venue":false,"about_ca":true,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":["sts","scholarly_communication","open_science"],"consensus_categories":[],"category_scores_codex":[0.005894681,0.00008598957,0.0001416135,0.00008114128,0.001321557,0.004202702,0.007472658,0.00003463063,0.0008108633],"category_scores_gemma":[0.0008856926,0.00006705217,0.00001298885,0.0006386616,0.000561861,0.001192918,0.001919586,0.0001233039,0.00002288601],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001723134,"about_ca_system_score_gemma":0.005017838,"about_ca_topic_candidate":true,"about_ca_topic_consensus":true,"about_ca_topic_score_codex":0.9967331,"about_ca_topic_score_gemma":0.9999045,"domain_scores_codex":[0.9977898,0.0002798308,0.0002205643,0.0009271619,0.0005109115,0.0002717313],"domain_scores_gemma":[0.9960254,0.0006988419,0.00003669285,0.003074318,0.00003136186,0.0001333649],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"not_applicable","study_design_gemma":"not_applicable","study_design_scores_codex":[0.000004941337,0.00002011406,0.001240505,0.000008473719,0.00002194424,0.000003797333,0.001105957,3.532922e-7,0.000002016558,0.002952039,0.9733344,0.02130543],"study_design_scores_gemma":[0.00009730473,0.000003769374,0.00325099,0.00002057264,0.00007003261,1.64152e-7,0.005849404,0.001842948,0.000001149179,0.003621983,0.9851487,0.00009295443],"study_design_candidate":"not_applicable","study_design_consensus":"not_applicable","genre_codex":"dataset","genre_gemma":"dataset","genre_scores_codex":[0.01980757,0.001573659,0.002934053,0.03203674,0.002039767,0.001239804,0.9385129,0.0000217539,0.001833754],"genre_scores_gemma":[0.483905,0.00002674865,0.001127274,0.001190181,0.000633481,0.00002268915,0.5104247,0.00001082268,0.002659053],"genre_candidate":"dataset","genre_consensus":"dataset","teacher_disagreement_score":0.4640975,"threshold_uncertainty_score":0.9999786,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1611647706221825,"score_gpt":0.4182198925424139,"score_spread":0.2570551219202314,"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."}}