{"id":"W4385571031","doi":"10.18653/v1/2023.findings-acl.868","title":"NusaCrowd: Open Source Initiative for Indonesian NLP Resources","year":2023,"lang":"en","type":"article","venue":"","topic":"Topic Modeling","field":"Computer Science","cited_by":32,"is_retracted":false,"has_abstract":true,"ca_institutions":"Kootenay Association for Science & Technology","funders":"","keywords":"Indonesian; Art; Artificial intelligence; Art history; Theology; Philosophy; Computer science; Linguistics","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":["open_science"],"consensus_categories":[],"category_scores_codex":[0.006742377,0.002169979,0.001933347,0.007826923,0.002781852,0.006489767,0.004298312,0.002092349,0.04505885],"category_scores_gemma":[0.02326572,0.002027586,0.002596843,0.007533232,0.001914437,0.01399493,0.01334691,0.004648359,0.04841894],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.002039939,"about_ca_system_score_gemma":0.006416656,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01127772,"about_ca_topic_score_gemma":0.01156361,"domain_scores_codex":[0.9951799,0.00128454,0.0007690659,0.001079591,0.001427744,0.0002591746],"domain_scores_gemma":[0.9836296,0.007251895,0.0007657227,0.004716051,0.001843272,0.001793401],"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":[0.0005679897,0.0001842422,0.001960602,0.002970081,0.0002160305,0.0007295643,0.00196995,0.001815286,0.004217893,0.01729319,0.8133937,0.1546815],"study_design_scores_gemma":[0.0001217278,0.00002830746,0.002554084,0.0004850425,0.0000569326,0.0003900977,0.0007850651,0.01174134,0.00452856,0.01926404,0.9598939,0.0001508739],"study_design_candidate":"not_applicable","study_design_consensus":"not_applicable","genre_codex":"software","genre_gemma":"dataset","genre_scores_codex":[0.006560793,0.002911272,0.2059348,0.005405745,0.001934023,0.001021081,0.304269,0.3949553,0.0770079],"genre_scores_gemma":[0.0314509,0.001876348,0.2944326,0.001493211,0.0002797028,0.002440843,0.544182,0.09318957,0.03065482],"genre_candidate":"dataset","genre_consensus":null,"teacher_disagreement_score":0.9957017,"threshold_uncertainty_score":0.1507369,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.08243536302076786,"score_gpt":0.3138575182029832,"score_spread":0.2314221551822154,"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."}}