{"meta":{"query_hash":"c357e742c7f2","filters":{"venue":"Journal of Open Humanities Data"},"cohort_total":9,"direct_labels_cover":0,"predictions_cover":9,"exported":9,"export_cap":100000,"truncated":false,"label_status":"direct model label, unvalidated","prediction_status":"machine_predicted_unvalidated (Codex and Gemma teacher distillation)","score_status":"score_only:v0-immature-baseline","snapshot":{"source":"OpenAlex, pinned release, all 482 partitions","release":"2026-06-24","frame_built":"2026-07-12"},"permalink":"https://metacan.xera.ac/q/c357e742c7f2","api":"https://metacan.xera.ac/api/v1/cohort?venue=Journal+of+Open+Humanities+Data"},"results":[{"id":"W3190643984","doi":"10.5334/johd.42","title":"The Corpus for Idiolectal Research (CIDRE)","year":2021,"lang":"en","type":"article","venue":"Journal of Open Humanities Data","topic":"Authorship Attribution and Profiling","field":"Computer Science","cited_by":1,"is_retracted":false,"has_abstract":true,"route_ca_aff":false,"route_ca_fund":false,"route_ca_venue":false,"route_about_ca":true,"ca_institutions":"","funders":"Agence Nationale de la Recherche","keywords":"Paratext; Scripting language; Metadata; License; Computer science; World Wide Web; Open source; Style (visual arts); Literature; Linguistics; Art; Programming language; Software; Philosophy","score_opus":0.6364291660620014,"score_gpt":0.5070253019086151,"score_spread":0.12940386415338634,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W3190643984","genre_codex":"methods","genre_gemma":"empirical","domain_codex":null,"domain_gemma":null,"model_version":"codex-gemma-dda1882f352a","genre_candidate":"methods","genre_consensus":null,"domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.009849429,0.015022938,0.89703554,0.038020946,0.0073336917,0.0012761494,0.00046151504,0.000054147218,0.030945646],"genre_scores_gemma":[0.734016,0.0035787225,0.19303228,0.0043248087,0.005703095,0.000059743896,0.0004585026,0.000098982,0.058727857],"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"not_applicable","domain_scores_codex":[0.9984014,0.00036262453,0.00035857473,0.00017555222,0.00044310885,0.00025872936],"domain_scores_gemma":[0.9967007,0.0010087274,0.00020393294,0.0009718923,0.0010570594,0.00005765899],"candidate_categories":["scholarly_communication","open_science"],"consensus_categories":[],"category_scores_codex":[0.00602987,0.000061031897,0.00014495286,0.000055255496,0.0010220897,0.002854078,0.0070238956,0.000029851655,0.00005150675],"category_scores_gemma":[0.0011270011,0.00004200391,0.000045964192,0.00020506607,0.00009743274,0.0017588586,0.003619271,0.00037372205,0.000013431881],"study_design_candidate":"not_applicable","study_design_consensus":null,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_system_candidate":false,"about_ca_system_consensus":false,"study_design_scores_codex":[0.00006884465,0.00004797125,0.00007516629,0.000016265049,0.00007142233,0.000112601396,0.00090329506,0.0000035704868,0.00038502977,0.7664054,0.2042795,0.027630912],"study_design_scores_gemma":[0.00038540384,0.00014590618,0.000104427774,0.000050910843,0.00000702195,0.00024745386,0.0011440505,0.0021780767,0.0016809666,0.02973528,0.96424145,0.00007907762],"about_ca_topic_score_codex":0.000017205548,"about_ca_topic_score_gemma":0.00010983241,"teacher_disagreement_score":0.7599619,"about_ca_system_score_codex":0.000050030154,"about_ca_system_score_gemma":0.0006641811,"threshold_uncertainty_score":0.9983486},"labels":[],"label_agreement":null},{"id":"W4308571509","doi":"10.5334/johd.97","title":"CREMMA Medii Aevi: Literary Manuscript Text Recognition in Latin","year":2023,"lang":"en","type":"article","venue":"Journal of Open Humanities Data","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":4,"is_retracted":false,"has_abstract":true,"route_ca_aff":true,"route_ca_fund":false,"route_ca_venue":false,"route_about_ca":false,"ca_institutions":"Université de Montréal","funders":"","keywords":"Latin Americans; Computer science; Representation (politics); Natural language processing; Artificial intelligence; Segmentation; Medieval Latin; Information retrieval; Linguistics; History; Classics; Political science","score_opus":0.23807547872461002,"score_gpt":0.3631402314952308,"score_spread":0.12506475277062076,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W4308571509","genre_codex":"empirical","genre_gemma":"methods","domain_codex":null,"domain_gemma":null,"model_version":"codex-gemma-dda1882f352a","genre_candidate":"methods","genre_consensus":null,"domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.44176063,0.09530307,0.23114663,0.07341314,0.023541506,0.007755398,0.002757518,0.0053729382,0.11894918],"genre_scores_gemma":[0.21707428,0.0011070546,0.77066594,0.0040486534,0.0012006925,0.000024469384,0.0006487126,0.00007700981,0.005153174],"study_design_codex":"design_other","study_design_gemma":"theoretical_or_conceptual","domain_scores_codex":[0.9984253,0.00013189482,0.0005718817,0.000251868,0.00039715672,0.00022186482],"domain_scores_gemma":[0.998445,0.00015052882,0.00039548654,0.00078260544,0.00017744683,0.000048930906],"candidate_categories":["scholarly_communication","open_science"],"consensus_categories":[],"category_scores_codex":[0.001623098,0.0001256937,0.00025810063,0.0005204928,0.000107436004,0.001511174,0.0066732536,0.000056144607,0.00012834236],"category_scores_gemma":[0.0003150031,0.00010590433,0.000032542368,0.0005205374,0.000045537006,0.0075490586,0.0035115604,0.00041485278,0.000059192826],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":null,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_system_candidate":false,"about_ca_system_consensus":false,"study_design_scores_codex":[0.00006957789,0.0001396528,0.00045493085,0.00015567822,0.000049662183,0.0021832814,0.0070192735,0.0000012689594,0.0010324077,0.011245975,0.45709655,0.52055174],"study_design_scores_gemma":[0.002246122,0.0006883827,0.0047636703,0.004090684,0.00005035517,0.0009240578,0.0018470951,0.009224102,0.005362446,0.77342016,0.19623797,0.0011449503],"about_ca_topic_score_codex":0.00009399636,"about_ca_topic_score_gemma":0.000105244544,"teacher_disagreement_score":0.7621742,"about_ca_system_score_codex":0.000060921477,"about_ca_system_score_gemma":0.000118873024,"threshold_uncertainty_score":0.99952537},"labels":[],"label_agreement":null},{"id":"W4312215389","doi":"10.5334/johd.94","title":"The TRANSCOMP Dataset of Literary Translations from 120 Languages and a Parallel Collection of English-language Originals","year":2022,"lang":"en","type":"article","venue":"Journal of Open Humanities Data","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":2,"is_retracted":false,"has_abstract":true,"route_ca_aff":true,"route_ca_fund":false,"route_ca_venue":false,"route_about_ca":false,"ca_institutions":"McGill University","funders":"","keywords":"Linguistics; Computer science; Word (group theory); Metadata; Literary translation; Natural language processing; Literature; History; Art; World Wide Web; Philosophy","score_opus":0.05700619531125964,"score_gpt":0.3531706120189455,"score_spread":0.2961644167076859,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W4312215389","genre_codex":"review","genre_gemma":"empirical","domain_codex":null,"domain_gemma":null,"model_version":"codex-gemma-dda1882f352a","genre_candidate":"empirical","genre_consensus":null,"domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.24136202,0.3859849,0.24630077,0.0051119886,0.0032054177,0.003408234,0.11095637,0.00029509774,0.003375209],"genre_scores_gemma":[0.665126,0.00078414887,0.3312152,0.00028212828,0.00026735588,0.000019963465,0.0020398789,0.000023550292,0.00024173898],"study_design_codex":"qualitative","study_design_gemma":"not_applicable","domain_scores_codex":[0.9986591,0.00021867189,0.00050980406,0.00016210508,0.00035081455,0.00009954373],"domain_scores_gemma":[0.99819946,0.0003571958,0.0005433396,0.00072269666,0.00015118346,0.000026100492],"candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0009834499,0.00009134463,0.00024285087,0.00012704634,0.00032172407,0.00037188633,0.0041258386,0.000019668483,0.00008256155],"category_scores_gemma":[0.00010634072,0.00006684343,0.000032681473,0.00020325145,0.000098051954,0.0020813094,0.000981051,0.00027287516,8.67767e-8],"study_design_candidate":"not_applicable","study_design_consensus":null,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_system_candidate":false,"about_ca_system_consensus":false,"study_design_scores_codex":[0.0015357194,0.0011924548,0.0011784182,0.00047910155,0.0012883091,0.00046203224,0.44659835,0.000099433775,0.015932124,0.09410277,0.3038163,0.133315],"study_design_scores_gemma":[0.0070276842,0.0030772567,0.0011878146,0.0010331693,0.00065115956,0.0005451445,0.0622686,0.011012456,0.021200456,0.08971287,0.8008808,0.0014025671],"about_ca_topic_score_codex":0.00044893182,"about_ca_topic_score_gemma":0.0001479819,"teacher_disagreement_score":0.4970645,"about_ca_system_score_codex":0.000020220961,"about_ca_system_score_gemma":0.00013176311,"threshold_uncertainty_score":0.7666903},"labels":[],"label_agreement":null},{"id":"W4312706369","doi":"10.5334/johd.88","title":"The CONLIT Dataset of Contemporary Literature","year":2022,"lang":"en","type":"article","venue":"Journal of Open Humanities Data","topic":"Narrative Theory and Analysis","field":"Arts and Humanities","cited_by":4,"is_retracted":false,"has_abstract":true,"route_ca_aff":true,"route_ca_fund":false,"route_ca_venue":false,"route_about_ca":false,"ca_institutions":"McGill University","funders":"","keywords":"Elite; Biography; Literature; History; Art","score_opus":0.2284273115647129,"score_gpt":0.3408972242918083,"score_spread":0.11246991272709539,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W4312706369","genre_codex":"other","genre_gemma":"empirical","domain_codex":null,"domain_gemma":null,"model_version":"codex-gemma-dda1882f352a","genre_candidate":"empirical","genre_consensus":null,"domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.054550074,0.055703096,0.000040436676,0.0101383645,0.0053668143,0.0013912218,0.40417218,0.000032433803,0.46860537],"genre_scores_gemma":[0.9576027,0.00026380937,0.000059402613,0.0013316885,0.0010957064,0.000008877482,0.010478606,0.000026242531,0.029132998],"study_design_codex":"not_applicable","study_design_gemma":"not_applicable","domain_scores_codex":[0.99829686,0.00044486954,0.0006705691,0.00012161417,0.00035223275,0.000113873066],"domain_scores_gemma":[0.9977241,0.0003122406,0.00090994994,0.0007849648,0.00023860873,0.000030144342],"candidate_categories":["sts","scholarly_communication","insufficient_payload"],"consensus_categories":[],"category_scores_codex":[0.002374055,0.00010899955,0.00032230417,0.00010405224,0.0015415117,0.0011104468,0.0041456027,0.0000100482075,0.007172388],"category_scores_gemma":[0.00008204076,0.00006741536,0.00007749601,0.00005713921,0.00046664,0.0019722495,0.0020572855,0.00038407507,0.0000046808505],"study_design_candidate":"not_applicable","study_design_consensus":"not_applicable","about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_system_candidate":false,"about_ca_system_consensus":false,"study_design_scores_codex":[0.00022791862,0.00008471308,0.000015291842,0.000013578593,0.00027461367,0.00004771155,0.016588558,0.0000023958025,0.00001635944,0.35402104,0.62846535,0.00024244915],"study_design_scores_gemma":[0.00035400072,0.00021717233,0.0000042228016,0.000037826172,0.00006060473,0.000032290576,0.070986696,0.0000126974965,0.000013712866,0.008195461,0.92000395,0.0000813485],"about_ca_topic_score_codex":0.00009311567,"about_ca_topic_score_gemma":0.00019405216,"teacher_disagreement_score":0.90305257,"about_ca_system_score_codex":0.000020928033,"about_ca_system_score_gemma":0.00014649473,"threshold_uncertainty_score":0.9999265},"labels":[],"label_agreement":null},{"id":"W4319597719","doi":"10.5334/johd.95","title":"MultiHATHI: A Complete Collection of Multilingual Prose Fiction in the HathiTrust Digital Library","year":2023,"lang":"en","type":"article","venue":"Journal of Open Humanities Data","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":2,"is_retracted":false,"has_abstract":true,"route_ca_aff":true,"route_ca_fund":false,"route_ca_venue":false,"route_about_ca":false,"ca_institutions":"McGill University","funders":"","keywords":"Bespoke; Metadata; Computer science; Digital library; Classifier (UML); English language; World Wide Web; Information retrieval; Natural language processing; Artificial intelligence; Linguistics; Literature; Art; Poetry","score_opus":0.13862741170011722,"score_gpt":0.3461929316426599,"score_spread":0.2075655199425427,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W4319597719","genre_codex":"empirical","genre_gemma":"empirical","domain_codex":null,"domain_gemma":null,"model_version":"codex-gemma-dda1882f352a","genre_candidate":"empirical","genre_consensus":"empirical","domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.8269555,0.012400227,0.09231337,0.015485459,0.005837437,0.008461528,0.0045318985,0.0021976703,0.031816915],"genre_scores_gemma":[0.92864996,0.000098731005,0.06989359,0.00026779957,0.0001913635,0.000008718805,0.00016702809,0.000018900997,0.00070389744],"study_design_codex":"not_applicable","study_design_gemma":"not_applicable","domain_scores_codex":[0.9987667,0.00010448287,0.00046443826,0.0001698661,0.00035832962,0.00013615514],"domain_scores_gemma":[0.99860275,0.000227359,0.00044014447,0.00060621375,0.00010477255,0.000018750055],"candidate_categories":["scholarly_communication"],"consensus_categories":[],"category_scores_codex":[0.0009355603,0.000094794086,0.0001961448,0.00031227164,0.00014075858,0.0015533703,0.005260484,0.00004035061,0.000015847463],"category_scores_gemma":[0.00028381307,0.00006512127,0.000035208937,0.00061980233,0.00007215431,0.007569914,0.0020628774,0.00031401857,0.000004577105],"study_design_candidate":"not_applicable","study_design_consensus":"not_applicable","about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_system_candidate":false,"about_ca_system_consensus":false,"study_design_scores_codex":[0.0008370737,0.001578193,0.006042131,0.0006639972,0.00024722828,0.0015019372,0.09623351,0.000117621625,0.0038621456,0.033881903,0.70281017,0.1522241],"study_design_scores_gemma":[0.012568569,0.0052996315,0.031686112,0.006282806,0.00020914659,0.0050733355,0.04883227,0.2584504,0.029697422,0.1609772,0.43787554,0.0030475662],"about_ca_topic_score_codex":0.000047816327,"about_ca_topic_score_gemma":0.000021719945,"teacher_disagreement_score":0.2649346,"about_ca_system_score_codex":0.000024931105,"about_ca_system_score_gemma":0.00014636577,"threshold_uncertainty_score":0.9994831},"labels":[],"label_agreement":null},{"id":"W4382727197","doi":"10.5334/johd.109","title":"A Mixed Method Twitter Methodology and Anonymous Corpus","year":2023,"lang":"en","type":"article","venue":"Journal of Open Humanities Data","topic":"Misinformation and Its Impacts","field":"Social Sciences","cited_by":0,"is_retracted":false,"has_abstract":true,"route_ca_aff":true,"route_ca_fund":false,"route_ca_venue":false,"route_about_ca":false,"ca_institutions":"Brock University","funders":"","keywords":"Vetting; Anonymity; Computer science; Field (mathematics); Work (physics); Process (computing); World Wide Web; Simple (philosophy); Data science; Internet privacy; Information retrieval; Computer security; Engineering; Mathematics","score_opus":0.7209269913846231,"score_gpt":0.5227081836891851,"score_spread":0.19821880769543798,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W4382727197","genre_codex":"other","genre_gemma":"methods","domain_codex":null,"domain_gemma":null,"model_version":"codex-gemma-dda1882f352a","genre_candidate":"other","genre_consensus":null,"domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.38812855,0.0012347244,0.020484935,0.028205244,0.0071359477,0.0014328492,0.0006048884,0.00017637823,0.5525965],"genre_scores_gemma":[0.30393595,0.009918678,0.4442974,0.031317387,0.0057520694,0.000008179623,0.0005691686,0.00014921362,0.20405196],"study_design_codex":"not_applicable","study_design_gemma":"not_applicable","domain_scores_codex":[0.9983313,0.0007067525,0.00036295858,0.000088753244,0.00030759777,0.00020262535],"domain_scores_gemma":[0.99844915,0.0006811149,0.00035241214,0.00026788661,0.00014526251,0.000104204424],"candidate_categories":["insufficient_payload"],"consensus_categories":[],"category_scores_codex":[0.00778106,0.00006875711,0.0002439497,0.00019712547,0.00037950635,0.00075193617,0.0014251142,0.000050552288,0.0010202054],"category_scores_gemma":[0.0016227245,0.000055900633,0.000024154882,0.00018055779,0.00015514284,0.0023340746,0.00074544153,0.0001564701,0.000047359805],"study_design_candidate":"not_applicable","study_design_consensus":"not_applicable","about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_system_candidate":false,"about_ca_system_consensus":false,"study_design_scores_codex":[0.00010329507,0.000024463508,0.00013194652,0.000022098357,0.00009834507,0.00006142068,0.1208106,0.000004140831,0.00008980297,0.046385717,0.7851202,0.047147963],"study_design_scores_gemma":[0.000516592,0.00007914054,0.0024863107,0.000027842692,0.000034369114,0.000090916124,0.0927248,0.000088043904,0.000040704566,0.0052401214,0.8985665,0.000104656254],"about_ca_topic_score_codex":0.0007405187,"about_ca_topic_score_gemma":0.00070503895,"teacher_disagreement_score":0.42381248,"about_ca_system_score_codex":0.000029523251,"about_ca_system_score_gemma":0.0002378324,"threshold_uncertainty_score":0.999893},"labels":[],"label_agreement":null},{"id":"W4389949574","doi":"10.5334/johd.169","title":"3D Modelling of Jerusalem’s Maghrebi Quarter","year":2023,"lang":"en","type":"article","venue":"Journal of Open Humanities Data","topic":"3D Surveying and Cultural Heritage","field":"Earth and Planetary Sciences","cited_by":1,"is_retracted":false,"has_abstract":true,"route_ca_aff":false,"route_ca_fund":false,"route_ca_venue":false,"route_about_ca":true,"ca_institutions":"","funders":"Université Gustave Eiffel; Università Degli Studi di Modena e Reggio Emila","keywords":"Quarter (Canadian coin); Geology; Archaeology; History","score_opus":0.3692056332634672,"score_gpt":0.3160634521328087,"score_spread":0.05314218113065855,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W4389949574","genre_codex":"empirical","genre_gemma":"empirical","domain_codex":null,"domain_gemma":null,"model_version":"codex-gemma-dda1882f352a","genre_candidate":"empirical","genre_consensus":"empirical","domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.9474395,0.0013032111,0.00039915578,0.00032368393,0.0011057876,0.00019069054,0.0019772435,0.000029970437,0.047230784],"genre_scores_gemma":[0.99486303,0.00032198403,0.0017144505,0.00008773471,0.00021313786,1.1827227e-7,0.0005175807,0.000004456461,0.002277517],"study_design_codex":"not_applicable","study_design_gemma":"not_applicable","domain_scores_codex":[0.99891067,0.000092115064,0.0004156135,0.00011672626,0.00029442893,0.00017042964],"domain_scores_gemma":[0.9991246,0.00012968965,0.00025655667,0.000341276,0.0000965912,0.000051323896],"candidate_categories":["insufficient_payload"],"consensus_categories":[],"category_scores_codex":[0.0012234359,0.00008486599,0.00023716436,0.000085889165,0.0001453022,0.00026159498,0.0017332238,0.000028802964,0.0029835692],"category_scores_gemma":[0.000029902498,0.00006030846,0.00003947813,0.00013465201,0.00006462668,0.0015696373,0.00012484744,0.00015349692,0.00016643891],"study_design_candidate":"not_applicable","study_design_consensus":"not_applicable","about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_system_candidate":false,"about_ca_system_consensus":false,"study_design_scores_codex":[0.0006849413,0.00014827965,0.18564677,0.00039538348,0.0005456427,0.0007237314,0.013391176,0.06301608,0.000121273326,0.0007327669,0.664722,0.069871925],"study_design_scores_gemma":[0.0029564665,0.0019351651,0.21040514,0.0010742898,0.00026063353,0.0004958281,0.048663292,0.26374787,0.00011157483,0.005984973,0.4630494,0.0013153331],"about_ca_topic_score_codex":0.0011103518,"about_ca_topic_score_gemma":0.0011800317,"teacher_disagreement_score":0.2016726,"about_ca_system_score_codex":0.000002366979,"about_ca_system_score_gemma":0.000052017578,"threshold_uncertainty_score":0.99792784},"labels":[],"label_agreement":null},{"id":"W4408373139","doi":"10.5334/johd.262","title":"Data from ARTiculating Canadian Identities: The Canada Council Art Bank’s Mycelial Presence and Influence in the Realm of Canadian Fine Art","year":2024,"lang":"en","type":"article","venue":"Journal of Open Humanities Data","topic":"Visual Culture and Art Theory","field":"Arts and Humanities","cited_by":0,"is_retracted":false,"has_abstract":true,"route_ca_aff":false,"route_ca_fund":true,"route_ca_venue":false,"route_about_ca":true,"ca_institutions":"","funders":"Social Sciences and Humanities Research Council of Canada","keywords":"Realm; Political science; Art; Public administration; Law","score_opus":0.22854600992274965,"score_gpt":0.30586384381278614,"score_spread":0.07731783389003649,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W4408373139","genre_codex":"empirical","genre_gemma":"empirical","domain_codex":null,"domain_gemma":null,"model_version":"codex-gemma-dda1882f352a","genre_candidate":"empirical","genre_consensus":"empirical","domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.613803,0.006619694,0.0000022273762,0.014369035,0.0026600992,0.0009694737,0.03854699,0.000012741859,0.32301673],"genre_scores_gemma":[0.9883953,0.00016328726,0.000014701152,0.0019716402,0.00085284025,0.000004016833,0.00057460414,0.000016955059,0.008006654],"study_design_codex":"not_applicable","study_design_gemma":"not_applicable","domain_scores_codex":[0.9981668,0.00015591673,0.0005743397,0.00022744048,0.0005866583,0.0002888174],"domain_scores_gemma":[0.997975,0.0004662169,0.00017969088,0.0009546434,0.0003031248,0.000121321966],"candidate_categories":["scholarly_communication","insufficient_payload"],"consensus_categories":[],"category_scores_codex":[0.0020489092,0.00015311108,0.0002523547,0.00015195055,0.0005661553,0.0023715277,0.003397843,0.000028180544,0.0020103573],"category_scores_gemma":[0.00054011797,0.00009065064,0.000021831636,0.00009863751,0.0003561103,0.0028532986,0.00058135594,0.00040511144,0.000018592718],"study_design_candidate":"not_applicable","study_design_consensus":"not_applicable","about_ca_topic_candidate":true,"about_ca_topic_consensus":true,"about_ca_system_candidate":false,"about_ca_system_consensus":false,"study_design_scores_codex":[0.000012642497,0.000008757732,0.0001778533,0.00003475139,0.00008253251,0.00028644523,0.02922783,0.000021044947,0.000008264256,0.033977367,0.93604684,0.00011567666],"study_design_scores_gemma":[0.00012432877,0.00003646476,0.0011126886,0.0003876414,0.000070425,0.000054385462,0.030497411,0.00029692787,0.0000019166391,0.0016534454,0.96564156,0.00012279222],"about_ca_topic_score_codex":0.996195,"about_ca_topic_score_gemma":0.999985,"teacher_disagreement_score":0.37459227,"about_ca_system_score_codex":0.00022456763,"about_ca_system_score_gemma":0.004988631,"threshold_uncertainty_score":0.99890196},"labels":[],"label_agreement":null},{"id":"W4416570569","doi":"10.5334/johd.383","title":"Digital Humanities Training and Workshops: Anonymized Survey Results","year":2025,"lang":"en","type":"article","venue":"Journal of Open Humanities Data","topic":"Digital Humanities and Scholarship","field":"Arts and Humanities","cited_by":0,"is_retracted":false,"has_abstract":true,"route_ca_aff":false,"route_ca_fund":true,"route_ca_venue":false,"route_about_ca":true,"ca_institutions":"","funders":"Social Sciences and Humanities Research Council of Canada; St. Francis Xavier University; University of Ottawa","keywords":"Certificate; Qualitative property; Digital humanities; Survey data collection; Digital media; Computer-assisted web interviewing; Qualitative research","score_opus":0.36401582024800605,"score_gpt":0.3469776897929208,"score_spread":0.017038130455085232,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W4416570569","genre_codex":"other","genre_gemma":"empirical","domain_codex":null,"domain_gemma":null,"model_version":"codex-gemma-dda1882f352a","genre_candidate":"other","genre_consensus":null,"domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.049082894,0.0018743583,0.000022946324,0.0003959169,0.0017215728,0.00038824778,0.0061052213,0.000048539863,0.9403603],"genre_scores_gemma":[0.78405493,0.00024813288,0.00011952067,0.0013061155,0.001014166,0.000006843146,0.0013364205,0.000055546498,0.21185829],"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"not_applicable","domain_scores_codex":[0.9970393,0.00018838211,0.0014184383,0.00042869084,0.00047282822,0.00045236057],"domain_scores_gemma":[0.9964712,0.001091616,0.0007362909,0.0009237064,0.000682722,0.000094425915],"candidate_categories":["metaepi_narrow","scholarly_communication"],"consensus_categories":[],"category_scores_codex":[0.0026143368,0.00037508857,0.0008443067,0.00047058662,0.0010515759,0.019914463,0.0036051513,0.000075357755,0.00091205566],"category_scores_gemma":[0.0010629715,0.0003201219,0.00011096647,0.000084056504,0.00096603925,0.011959349,0.0022885895,0.0005879591,0.00001556019],"study_design_candidate":"not_applicable","study_design_consensus":null,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_system_candidate":false,"about_ca_system_consensus":false,"study_design_scores_codex":[0.0022213953,0.00039308783,0.00040861111,0.00012721545,0.001074384,0.00021909543,0.054361414,0.000002977734,0.000002472169,0.550038,0.34633452,0.044816837],"study_design_scores_gemma":[0.0024538308,0.00031620343,0.0009914879,0.00043955527,0.00010519119,0.00005355598,0.06957553,0.00001717896,0.0000025519194,0.008608853,0.9170391,0.00039693995],"about_ca_topic_score_codex":0.00032461953,"about_ca_topic_score_gemma":0.005232507,"teacher_disagreement_score":0.73497206,"about_ca_system_score_codex":0.00007071863,"about_ca_system_score_gemma":0.00032086557,"threshold_uncertainty_score":0.9999251},"labels":[],"label_agreement":null}]}