{"id":"W4408403663","doi":"10.1145/3677389.3702566","title":"A Study on Building the World Literature Data Collection for Digital Humanities","year":2024,"lang":"en","type":"article","venue":"","topic":"Asian Culture and Media Studies","field":"Social Sciences","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Toronto","funders":"","keywords":"Digital humanities; Computer science; Data science; World Wide Web; Humanities; Art","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":[0.04504194,0.0003761379,0.0004505262,0.009807806,0.005443068,0.006971143,0.001617849,0.0006889182,0.002337549],"category_scores_gemma":[0.1011054,0.0006082356,0.0008614837,0.01664404,0.002549504,0.01037545,0.003928714,0.001426499,0.0006604607],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.008927696,"about_ca_system_score_gemma":0.01584814,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.02756178,"about_ca_topic_score_gemma":0.03387454,"domain_scores_codex":[0.971761,0.01728663,0.003208833,0.0017887,0.005048629,0.0009062364],"domain_scores_gemma":[0.9254128,0.04010016,0.005376753,0.007654562,0.01965639,0.001799284],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"qualitative","study_design_gemma":"not_applicable","study_design_scores_codex":[0.0002114107,0.0007501783,0.3114032,0.001682751,0.00006246098,0.001209199,0.3197951,0.000740063,0.003561823,0.06421518,0.01435274,0.2820159],"study_design_scores_gemma":[0.00005412759,0.0004154049,0.1313952,0.002288324,0.0001445039,0.0006911391,0.6490816,0.005221867,0.00477745,0.007599341,0.198209,0.0001219841],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"methods","genre_scores_codex":[0.8862234,0.0006945689,0.0471564,0.006396097,0.0001662612,0.006942765,0.003732726,0.0003026496,0.0483852],"genre_scores_gemma":[0.8555355,0.0009976876,0.1260994,0.0009059215,0.00003878495,0.005987793,0.005033072,0.0001594523,0.005242442],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.04504194,"threshold_uncertainty_score":0.2382075,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1194298653468102,"score_gpt":0.4004230180841166,"score_spread":0.2809931527373064,"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."}}