{"meta":{"query_hash":"151acead7183","filters":{"venue":"2022 IEEE International Conference on Big Data (Big Data)"},"cohort_total":41,"direct_labels_cover":0,"predictions_cover":41,"exported":41,"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/151acead7183","api":"https://metacan.xera.ac/api/v1/cohort?venue=2022+IEEE+International+Conference+on+Big+Data+%28Big+Data%29"},"results":[{"id":"W4226052253","doi":"10.1109/bigdata55660.2022.10020281","title":"PMFL: Partial Meta-Federated Learning for heterogeneous tasks and its applications on real-world medical records","year":2022,"lang":"en","type":"preprint","venue":"2022 IEEE International Conference on Big Data (Big Data)","topic":"Privacy-Preserving Technologies in Data","field":"Computer Science","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":"Mila - Quebec Artificial Intelligence Institute","funders":"Fundamental Research Funds for the Central Universities","keywords":"Computer science; Federated learning; Transfer of learning; Meta learning (computer science); Task (project management); Machine learning; Generalization; Artificial intelligence; Homogeneous; Exploit","score_opus":0.4329335683524456,"score_gpt":0.41144588545225624,"score_spread":0.021487682900189375,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W4226052253","genre_codex":"methods","genre_gemma":"methods","domain_codex":null,"domain_gemma":null,"model_version":"metacan-v3-hybrid-931329e0061c","genre_candidate":"methods","genre_consensus":"methods","domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.049036615,0.0005332097,0.9430329,0.0009310658,0.0000863206,0.000158156,0.00034894876,0.004984394,0.0008883804],"genre_scores_gemma":[0.5784202,0.00017282454,0.4171147,0.00053659006,0.000071781666,0.00026573736,0.0013266122,0.00022072239,0.0018708124],"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","domain_scores_codex":[0.9982469,0.0007234514,0.0001345198,0.00043791108,0.00026830638,0.00018890497],"domain_scores_gemma":[0.9965102,0.0014696708,0.00019025957,0.001144063,0.00049144006,0.00019430873],"candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0056173485,0.001002542,0.0011395128,0.0014969435,0.00095528894,0.0014139405,0.0026141154,0.0022603504,0.0016472513],"category_scores_gemma":[0.011024208,0.00041206094,0.0012360417,0.0013793724,0.00085474947,0.0031243437,0.0026204812,0.001456701,0.00052483665],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_system_candidate":false,"about_ca_system_consensus":false,"study_design_scores_codex":[0.00069330115,0.00048347536,0.006185242,0.00017513859,0.00022776326,0.00024192691,0.0002527941,0.5033939,0.002866019,0.0057034213,0.008904213,0.47087285],"study_design_scores_gemma":[0.000037797283,0.000059265207,0.00037620263,0.000012576172,0.0000168204,0.000084828185,0.000032467997,0.98864007,0.0017681685,0.008003837,0.0009586975,0.00000931082],"about_ca_topic_score_codex":0.0036805207,"about_ca_topic_score_gemma":0.004059264,"teacher_disagreement_score":0.0056173485,"about_ca_system_score_codex":0.0012274886,"about_ca_system_score_gemma":0.002009205,"threshold_uncertainty_score":0.02970773},"labels":[],"label_agreement":null},{"id":"W4317781640","doi":"10.1109/bigdata55660.2022.10020852","title":"The Lannion report on Big Data and Security Monitoring Research","year":2022,"lang":"en","type":"article","venue":"2022 IEEE International Conference on Big Data (Big Data)","topic":"Network Security and Intrusion Detection","field":"Computer Science","cited_by":1,"is_retracted":false,"has_abstract":true,"route_ca_aff":true,"route_ca_fund":false,"route_ca_venue":false,"route_about_ca":false,"ca_institutions":"Bell (Canada)","funders":"Engineering and Physical Sciences Research Council; Institut national de recherche en informatique et en automatique (INRIA)","keywords":"Big data; Computer science; Data science; Analytics; Data management; Computer security; Computation; Data mining","score_opus":0.46222809716017266,"score_gpt":0.41475548439871396,"score_spread":0.0474726127614587,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W4317781640","genre_codex":"commentary","genre_gemma":"other","domain_codex":null,"domain_gemma":null,"model_version":"metacan-v3-hybrid-931329e0061c","genre_candidate":"other","genre_consensus":null,"domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.0029376503,0.15694544,0.019033851,0.4100804,0.20651568,0.0008573867,0.0317159,0.00512211,0.16679154],"genre_scores_gemma":[0.030280504,0.2713797,0.036071964,0.13355702,0.12102702,0.0017098453,0.09679031,0.0035899484,0.3055937],"study_design_codex":"not_applicable","study_design_gemma":"not_applicable","domain_scores_codex":[0.98077786,0.0025883496,0.0010142258,0.0013612575,0.0129598975,0.0012983867],"domain_scores_gemma":[0.95017534,0.01562389,0.0026989197,0.0031299465,0.023481496,0.00489043],"candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.020429041,0.003071628,0.0011947743,0.011443351,0.001913456,0.009701525,0.0021772254,0.00519304,0.019253392],"category_scores_gemma":[0.031417817,0.000784034,0.0016839348,0.011807354,0.0014768997,0.011232157,0.005483309,0.006553189,0.016574709],"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":[0.000051249364,0.000028759421,0.000581548,0.00024033531,0.00002534145,0.00004621841,0.000038898917,0.00017294814,0.00022929283,0.0034129736,0.9642434,0.03092918],"study_design_scores_gemma":[0.000010951853,0.000022974718,0.0010735857,0.00043392382,0.000027117447,0.000034787896,0.000078527344,0.00028822842,0.0005852003,0.0017894726,0.9956332,0.000022021994],"about_ca_topic_score_codex":0.018462986,"about_ca_topic_score_gemma":0.019784665,"teacher_disagreement_score":0.020429041,"about_ca_system_score_codex":0.005979531,"about_ca_system_score_gemma":0.011862317,"threshold_uncertainty_score":0.10804039},"labels":[],"label_agreement":null},{"id":"W4318147159","doi":"10.1109/bigdata55660.2022.10020959","title":"Adaptive Content Forwarding Mechanism for Platoon based Vehicular Named Data Networks","year":2022,"lang":"en","type":"article","venue":"2022 IEEE International Conference on Big Data (Big Data)","topic":"Caching and Content Delivery","field":"Computer Science","cited_by":11,"is_retracted":false,"has_abstract":true,"route_ca_aff":true,"route_ca_fund":false,"route_ca_venue":false,"route_about_ca":false,"ca_institutions":"Brandon University","funders":"","keywords":"Platoon; Computer science; Computer network; Overhead (engineering); Scheme (mathematics); Throughput; The Internet; Vehicular ad hoc network; Data exchange; Electronic data interchange; Protocol (science); Distributed computing; Wireless ad hoc network; Telecommunications; Control (management); Wireless; Database; Operating system","score_opus":0.47588963696264636,"score_gpt":0.33149483326991913,"score_spread":0.14439480369272722,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W4318147159","genre_codex":"methods","genre_gemma":"empirical","domain_codex":null,"domain_gemma":null,"model_version":"metacan-v3-hybrid-931329e0061c","genre_candidate":"empirical","genre_consensus":null,"domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.11240073,0.00077538716,0.87643963,0.0004357568,0.00037650912,0.00035377685,0.00025914607,0.0023840335,0.006575055],"genre_scores_gemma":[0.9280716,0.00029478836,0.06892936,0.00010077175,0.000048990045,0.000114494454,0.0003419272,0.000034478864,0.0020635182],"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","domain_scores_codex":[0.99959356,0.00008512679,0.000031477066,0.00008297485,0.0001288791,0.000078008416],"domain_scores_gemma":[0.9989766,0.0002747771,0.00013138712,0.00018845427,0.0003422638,0.00008654725],"candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.00085181434,0.00039542493,0.0004712314,0.0010841334,0.0009482296,0.00058394537,0.0014345618,0.00053440145,0.00096633844],"category_scores_gemma":[0.0018426724,0.0001698763,0.00034308716,0.00070135517,0.00044525365,0.0011269973,0.00084207684,0.00037699766,0.00024616194],"study_design_candidate":"simulation_or_modeling","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":[0.0017488736,0.0004955978,0.010505963,0.00042335657,0.00028547508,0.001470641,0.0014171004,0.32778275,0.1272301,0.12455093,0.019500209,0.38458902],"study_design_scores_gemma":[0.000051636882,0.0001432011,0.00074042956,0.000012530566,0.000049005703,0.00023858304,0.00009561182,0.96708995,0.013054249,0.009558567,0.008929804,0.00003645568],"about_ca_topic_score_codex":0.005108405,"about_ca_topic_score_gemma":0.004444557,"teacher_disagreement_score":0.005108405,"about_ca_system_score_codex":0.0011654173,"about_ca_system_score_gemma":0.0009251614,"threshold_uncertainty_score":0.010157347},"labels":[],"label_agreement":null},{"id":"W4318147167","doi":"10.1109/bigdata55660.2022.10020867","title":"Emergence of an Autonomous Vehicle Secondary Data Market for Breakthrough Applications","year":2022,"lang":"en","type":"article","venue":"2022 IEEE International Conference on Big Data (Big Data)","topic":"Human Mobility and Location-Based Analysis","field":"Social Sciences","cited_by":5,"is_retracted":false,"has_abstract":true,"route_ca_aff":true,"route_ca_fund":false,"route_ca_venue":false,"route_about_ca":false,"ca_institutions":"Carleton University","funders":"","keywords":"Computer science; Data science; Cloud computing; Variety (cybernetics); Architecture; Data collection; Key (lock); Telecommunications; Systems engineering; Artificial intelligence; Engineering; Computer security; Geography","score_opus":0.3288503865665113,"score_gpt":0.4098341532725872,"score_spread":0.08098376670607593,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W4318147167","genre_codex":"methods","genre_gemma":"empirical","domain_codex":null,"domain_gemma":null,"model_version":"metacan-v3-hybrid-931329e0061c","genre_candidate":"empirical","genre_consensus":null,"domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.19145735,0.004743397,0.53159463,0.13458148,0.0028587223,0.0006190997,0.0029234982,0.0043239603,0.1268979],"genre_scores_gemma":[0.81975734,0.0039072847,0.13960943,0.0061817,0.0016975716,0.00028426654,0.002300225,0.0005847424,0.025677457],"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"theoretical_or_conceptual","domain_scores_codex":[0.9984611,0.00028798776,0.00005704956,0.00028189446,0.0006452512,0.00026674164],"domain_scores_gemma":[0.9925458,0.0017796246,0.00041529888,0.0011555769,0.0029163873,0.0011873071],"candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0046963124,0.0004138353,0.0004908085,0.0010110924,0.0011272174,0.006888398,0.002587081,0.0027085082,0.0077804327],"category_scores_gemma":[0.007741979,0.00033239948,0.00059351267,0.00162238,0.001856261,0.0136618465,0.0040309518,0.0032689075,0.0025156946],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":"theoretical_or_conceptual","about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_system_candidate":false,"about_ca_system_consensus":false,"study_design_scores_codex":[0.00027938833,0.00022576189,0.004929757,0.00028143855,0.00003379345,0.00049716,0.0008591732,0.008742684,0.009613982,0.79304326,0.057420596,0.12407297],"study_design_scores_gemma":[0.00007628367,0.00044376304,0.003502334,0.00021942821,0.000031845502,0.00051616045,0.0023862852,0.21455146,0.009551401,0.3282207,0.44039986,0.00010039787],"about_ca_topic_score_codex":0.002788791,"about_ca_topic_score_gemma":0.0021071264,"teacher_disagreement_score":0.0077804327,"about_ca_system_score_codex":0.002729588,"about_ca_system_score_gemma":0.0027479043,"threshold_uncertainty_score":0.026028156},"labels":[],"label_agreement":null},{"id":"W4318147168","doi":"10.1109/bigdata55660.2022.10020831","title":"A Real-time, Scalable Monitoring and User Analytics Solution for Microservices-based Software Applications","year":2022,"lang":"en","type":"article","venue":"2022 IEEE International Conference on Big Data (Big Data)","topic":"Software System Performance and Reliability","field":"Computer Science","cited_by":3,"is_retracted":false,"has_abstract":true,"route_ca_aff":true,"route_ca_fund":true,"route_ca_venue":false,"route_about_ca":false,"ca_institutions":"Carleton University","funders":"Natural Sciences and Engineering Research Council of Canada","keywords":"Microservices; Computer science; Scalability; Analytics; Big data; Database transaction; Transaction log; Software; Distributed computing; Database; Cloud computing; Data mining; Operating system","score_opus":0.18264378705047773,"score_gpt":0.34405238847640135,"score_spread":0.16140860142592361,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W4318147168","genre_codex":"methods","genre_gemma":"empirical","domain_codex":null,"domain_gemma":null,"model_version":"metacan-v3-hybrid-931329e0061c","genre_candidate":"empirical","genre_consensus":null,"domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.2114588,0.000554506,0.72736275,0.001404889,0.00013784552,0.00046839216,0.00056568475,0.05198727,0.006059853],"genre_scores_gemma":[0.7983428,0.00018683981,0.19498423,0.0003866736,0.00006108742,0.0002478119,0.0006208827,0.00040187797,0.004767688],"study_design_codex":"design_other","study_design_gemma":"bench_or_experimental","domain_scores_codex":[0.9990319,0.000114013375,0.00007846726,0.00026059596,0.00039454593,0.000120464625],"domain_scores_gemma":[0.9991585,0.00016774375,0.000120711746,0.00019728414,0.00021373043,0.00014209682],"candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.00071794254,0.0007325864,0.0005518082,0.0005647579,0.0004490183,0.0010786579,0.0015893559,0.00065441866,0.0011982779],"category_scores_gemma":[0.0015752586,0.00040582864,0.0003366014,0.0005104844,0.00032793864,0.0016365655,0.0014846708,0.0012703369,0.00064079626],"study_design_candidate":"bench_or_experimental","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":[0.0015194955,0.0012354223,0.026291596,0.0004669323,0.0001963845,0.0016764728,0.0016191204,0.049443524,0.17745131,0.012777491,0.023182092,0.70414007],"study_design_scores_gemma":[0.000058197173,0.00042038015,0.012202312,0.00007482078,0.00006572353,0.0004607217,0.0004220665,0.8970878,0.06447021,0.007778035,0.016849125,0.00011056698],"about_ca_topic_score_codex":0.0018633434,"about_ca_topic_score_gemma":0.002328838,"teacher_disagreement_score":0.0018633434,"about_ca_system_score_codex":0.0004480394,"about_ca_system_score_gemma":0.000876137,"threshold_uncertainty_score":0.0040087104},"labels":[],"label_agreement":null},{"id":"W4318147483","doi":"10.1109/bigdata55660.2022.10020802","title":"A Graph-based Analysis Approach to Cluster Lifetime Dynamics","year":2022,"lang":"en","type":"article","venue":"2022 IEEE International Conference on Big Data (Big Data)","topic":"Data Management and Algorithms","field":"Computer Science","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":"University of Waterloo","funders":"","keywords":"Computer science; Cluster analysis; Cluster (spacecraft); Graph; Data mining; Outlier; Power graph analysis; Theoretical computer science; Visualization; Data visualization; Artificial intelligence","score_opus":0.21032828965218048,"score_gpt":0.3246473363864822,"score_spread":0.11431904673430171,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W4318147483","genre_codex":"methods","genre_gemma":"empirical","domain_codex":null,"domain_gemma":null,"model_version":"metacan-v3-hybrid-931329e0061c","genre_candidate":"empirical","genre_consensus":null,"domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.0062007196,0.0004198831,0.9896271,0.00050509366,0.000044727294,0.00008770117,0.00056516693,0.00057454564,0.00197499],"genre_scores_gemma":[0.22752726,0.0017449899,0.7622585,0.00034072023,0.00018187556,0.00030238045,0.0021700084,0.00042574923,0.0050485083],"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","domain_scores_codex":[0.99910897,0.00026299647,0.000053900654,0.00027084566,0.00024260995,0.000060569313],"domain_scores_gemma":[0.9966828,0.0016306903,0.00041206798,0.00044284353,0.000713564,0.000117974254],"candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0011655294,0.0011531201,0.0007435842,0.006532225,0.0012116249,0.00272646,0.0014935242,0.00097558653,0.0027127166],"category_scores_gemma":[0.0057714377,0.00054498384,0.0013151497,0.006080729,0.0011047239,0.0038901896,0.001173108,0.0017632935,0.00065727456],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_system_candidate":false,"about_ca_system_consensus":false,"study_design_scores_codex":[0.00012811864,0.0001775587,0.008137002,0.0004987851,0.00034105073,0.0005536452,0.0016789035,0.41504097,0.010463484,0.34373567,0.013247969,0.20599686],"study_design_scores_gemma":[0.00001007027,0.0000391394,0.0017482233,0.00005193561,0.000059807742,0.00024835073,0.0002989972,0.8276446,0.0017794648,0.15192658,0.016146235,0.000046582667],"about_ca_topic_score_codex":0.012756425,"about_ca_topic_score_gemma":0.009784157,"teacher_disagreement_score":0.012756425,"about_ca_system_score_codex":0.00224258,"about_ca_system_score_gemma":0.0012404633,"threshold_uncertainty_score":0.025364399},"labels":[],"label_agreement":null},{"id":"W4318147587","doi":"10.1109/bigdata55660.2022.10021025","title":"Handwritten Word Recognition using Deep Learning Approach: A Novel Way of Generating Handwritten Words","year":2022,"lang":"en","type":"article","venue":"2022 IEEE International Conference on Big Data (Big Data)","topic":"Handwritten Text Recognition Techniques","field":"Computer Science","cited_by":11,"is_retracted":false,"has_abstract":true,"route_ca_aff":true,"route_ca_fund":false,"route_ca_venue":false,"route_about_ca":false,"ca_institutions":"University of Manitoba","funders":"","keywords":"Computer science; Word (group theory); Artificial intelligence; Word error rate; Bengali; Natural language processing; Process (computing); Speech recognition; Recall rate; Deep learning; Precision and recall; Language model; Pattern recognition (psychology); Mathematics","score_opus":0.3210273366955822,"score_gpt":0.3386155405561522,"score_spread":0.017588203860569973,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W4318147587","genre_codex":"methods","genre_gemma":"empirical","domain_codex":null,"domain_gemma":null,"model_version":"metacan-v3-hybrid-931329e0061c","genre_candidate":"empirical","genre_consensus":null,"domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.078565344,0.00053643575,0.9082318,0.0002518286,0.00019032242,0.00014602744,0.00039506922,0.0063925176,0.0052905697],"genre_scores_gemma":[0.5356234,0.00041839742,0.44856226,0.00035143434,0.00004950159,0.00014208048,0.0017124905,0.00031150682,0.012829005],"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","domain_scores_codex":[0.9996643,0.000042292933,0.00002597068,0.000116956755,0.000111729234,0.00003870727],"domain_scores_gemma":[0.9996165,0.000084582985,0.000036346788,0.000091414164,0.00014693338,0.00002414616],"candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0003108816,0.00060814637,0.00035661043,0.00042328794,0.00018673074,0.0005349625,0.000819704,0.00052439614,0.0017229322],"category_scores_gemma":[0.0007860306,0.00023433033,0.00049610523,0.00051091786,0.00030654832,0.0010125663,0.0005326018,0.00085586886,0.001048249],"study_design_candidate":"simulation_or_modeling","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":[0.00018874147,0.00017922114,0.0014925235,0.000194295,0.00006797798,0.00036023813,0.00013173878,0.075993925,0.11897529,0.0031448726,0.00528243,0.7939887],"study_design_scores_gemma":[0.000027607535,0.00021319446,0.0011789224,0.00002342152,0.000034242956,0.00041972098,0.00004093221,0.85885406,0.12798235,0.0033506968,0.007843582,0.00003122439],"about_ca_topic_score_codex":0.0024881351,"about_ca_topic_score_gemma":0.003555537,"teacher_disagreement_score":0.0024881351,"about_ca_system_score_codex":0.0004120735,"about_ca_system_score_gemma":0.0006035463,"threshold_uncertainty_score":0.005763769},"labels":[],"label_agreement":null},{"id":"W4318147647","doi":"10.1109/bigdata55660.2022.10020940","title":"Using CNN-LSTM Model for Weather Forecasting","year":2022,"lang":"en","type":"article","venue":"2022 IEEE International Conference on Big Data (Big Data)","topic":"Energy Load and Power Forecasting","field":"Engineering","cited_by":22,"is_retracted":false,"has_abstract":true,"route_ca_aff":true,"route_ca_fund":false,"route_ca_venue":false,"route_about_ca":false,"ca_institutions":"Carleton University","funders":"","keywords":"Computer science; Deep learning; Generalization; Artificial neural network; Weather forecasting; Artificial intelligence; Feature (linguistics); Data modeling; Big data; Machine learning; Numerical weather prediction; Data mining; Meteorology; Database; Geography","score_opus":0.5383621885771043,"score_gpt":0.3526381344140766,"score_spread":0.18572405416302767,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W4318147647","genre_codex":"methods","genre_gemma":"empirical","domain_codex":null,"domain_gemma":null,"model_version":"metacan-v3-hybrid-931329e0061c","genre_candidate":"empirical","genre_consensus":null,"domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.27139908,0.0034249616,0.6833476,0.0020328788,0.0012236506,0.00009863278,0.005194104,0.01221071,0.021068322],"genre_scores_gemma":[0.93825775,0.000852547,0.048059605,0.00025243626,0.0000958981,0.000056571356,0.0036619804,0.0001051743,0.008658162],"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","domain_scores_codex":[0.9999224,0.000007764602,0.0000051464294,0.000026239066,0.000018931798,0.000019547164],"domain_scores_gemma":[0.9999069,0.000026644539,0.000009183639,0.0000114404775,0.000040530973,0.0000052480323],"candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.00018006853,0.00065959763,0.00029516837,0.00026217708,0.00016374348,0.00045374353,0.00058789115,0.00057734863,0.0026649716],"category_scores_gemma":[0.0005333254,0.00022718565,0.000384063,0.00042864666,0.00011737556,0.00088732166,0.000279298,0.0008568015,0.0009582448],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_system_candidate":false,"about_ca_system_consensus":false,"study_design_scores_codex":[0.00022721266,0.00012323838,0.002657514,0.0001290707,0.00011384541,0.00016788188,0.000032352662,0.7532723,0.018478418,0.0019798013,0.010253845,0.21256454],"study_design_scores_gemma":[0.0000027462274,0.00001385205,0.00031007006,0.000004379395,0.000006458496,0.000009944314,0.0000034995655,0.9964779,0.0020937077,0.00061788474,0.00045561028,0.0000039245333],"about_ca_topic_score_codex":0.02104211,"about_ca_topic_score_gemma":0.024398293,"teacher_disagreement_score":0.02104211,"about_ca_system_score_codex":0.00053533644,"about_ca_system_score_gemma":0.00056344277,"threshold_uncertainty_score":0.041839242},"labels":[],"label_agreement":null},{"id":"W4318147653","doi":"10.1109/bigdata55660.2022.10020557","title":"Multimodal Deep Homography Estimation Using a Domain Adaptation Generative Adversarial Network","year":2022,"lang":"en","type":"article","venue":"2022 IEEE International Conference on Big Data (Big Data)","topic":"Advanced Image and Video Retrieval 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":"Defence Research and Development Canada; National Research Council Canada; Thales (Canada)","funders":"","keywords":"Artificial intelligence; Computer science; Homography; Computer vision; Parallax; Deep learning; Feature (linguistics); Generative grammar; Modalities; Domain (mathematical analysis); Adversarial system; Pattern recognition (psychology); Mathematics; Projective test","score_opus":0.27058870296074505,"score_gpt":0.3671188911458947,"score_spread":0.09653018818514963,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W4318147653","genre_codex":"methods","genre_gemma":"empirical","domain_codex":null,"domain_gemma":null,"model_version":"metacan-v3-hybrid-931329e0061c","genre_candidate":"empirical","genre_consensus":null,"domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.0112148,0.00029607158,0.98461956,0.00024595435,0.000052120868,0.000042524483,0.00011000986,0.0017983519,0.0016207262],"genre_scores_gemma":[0.62298536,0.00065254216,0.35536698,0.00087803614,0.00018694058,0.00021328997,0.00152367,0.0006675881,0.017525623],"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","domain_scores_codex":[0.9995747,0.000106489926,0.000013597823,0.00013970338,0.00010649907,0.0000589826],"domain_scores_gemma":[0.9995098,0.00019417249,0.000056346893,0.00012432279,0.00007580359,0.000039564013],"candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.00078424695,0.0012312868,0.001328496,0.0006141839,0.00033370894,0.0007927343,0.0017261024,0.0012930576,0.004018029],"category_scores_gemma":[0.0017318388,0.0006153238,0.0011747378,0.0006810514,0.00087337726,0.0013007933,0.0020023258,0.0028578462,0.0013987581],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_system_candidate":false,"about_ca_system_consensus":false,"study_design_scores_codex":[0.0001294847,0.000089191766,0.00058636116,0.000066231645,0.000112472735,0.00015196808,0.00006971301,0.79545784,0.0064146305,0.009274618,0.0058654523,0.18178198],"study_design_scores_gemma":[0.000003973404,0.000011970173,0.000059850972,0.0000039215884,0.0000044124286,0.000027492106,0.000004094208,0.9963257,0.0008384239,0.002246701,0.00046934513,0.0000041878557],"about_ca_topic_score_codex":0.005216438,"about_ca_topic_score_gemma":0.0069335774,"teacher_disagreement_score":0.005216438,"about_ca_system_score_codex":0.0008499872,"about_ca_system_score_gemma":0.0009685078,"threshold_uncertainty_score":0.013441682},"labels":[],"label_agreement":null},{"id":"W4318147664","doi":"10.1109/bigdata55660.2022.10021035","title":"Analysis of Airfare during Pandemic: A Multi-Agent Based Modeling Approach","year":2022,"lang":"en","type":"article","venue":"2022 IEEE International Conference on Big Data (Big Data)","topic":"Aviation Industry Analysis and Trends","field":"Economics, Econometrics and Finance","cited_by":1,"is_retracted":false,"has_abstract":true,"route_ca_aff":true,"route_ca_fund":false,"route_ca_venue":false,"route_about_ca":false,"ca_institutions":"Carleton University","funders":"","keywords":"Pandemic; Coronavirus disease 2019 (COVID-19); Air travel; Computer science; 2019-20 coronavirus outbreak; Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2); Aviation; Business; Risk analysis (engineering); Operations research; Engineering; Aerospace engineering; Infectious disease (medical specialty)","score_opus":0.5808910784080009,"score_gpt":0.3499821377569418,"score_spread":0.2309089406510591,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W4318147664","genre_codex":"methods","genre_gemma":"empirical","domain_codex":null,"domain_gemma":null,"model_version":"metacan-v3-hybrid-931329e0061c","genre_candidate":"empirical","genre_consensus":null,"domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.2414849,0.002592124,0.7102187,0.0049317824,0.00036310536,0.00035399754,0.0018979945,0.00042233823,0.037735026],"genre_scores_gemma":[0.9484584,0.0013836535,0.03903445,0.00021319358,0.00015235589,0.00029584207,0.00055481953,0.00005290829,0.009854268],"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","domain_scores_codex":[0.99962044,0.00016549323,0.000021826081,0.00007356124,0.000056449735,0.00006219599],"domain_scores_gemma":[0.99937326,0.00034279635,0.0001227886,0.000018577952,0.000087931665,0.00005464969],"candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0006742024,0.00083082414,0.0008777208,0.00086113485,0.00064572186,0.0015980242,0.0016237313,0.0021158012,0.002836693],"category_scores_gemma":[0.0013559994,0.00052062346,0.0013045588,0.00067147554,0.000490061,0.0013633735,0.0010446755,0.0012447579,0.00027877867],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_system_candidate":false,"about_ca_system_consensus":false,"study_design_scores_codex":[0.000014034906,0.000026916852,0.0011700507,0.000029284553,0.000042665495,0.00009187254,0.000044243283,0.9897367,0.0002665119,0.0067319605,0.00039715742,0.0014487404],"study_design_scores_gemma":[0.0000039970687,0.000011772094,0.00023269493,0.0000048345537,0.000009353374,0.000009606134,0.000025583617,0.99769056,0.00002719614,0.0015912165,0.00038798727,0.000005159083],"about_ca_topic_score_codex":0.028472856,"about_ca_topic_score_gemma":0.014383363,"teacher_disagreement_score":0.028472856,"about_ca_system_score_codex":0.0013119945,"about_ca_system_score_gemma":0.001370858,"threshold_uncertainty_score":0.05661422},"labels":[],"label_agreement":null},{"id":"W4318147681","doi":"10.1109/bigdata55660.2022.10020382","title":"A Framework for the Design, Development, Testing and Deployment of Reliable Big Data Platforms","year":2022,"lang":"en","type":"article","venue":"2022 IEEE International Conference on Big Data (Big Data)","topic":"Distributed systems and fault tolerance","field":"Computer Science","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":"Ontario Tech University","funders":"","keywords":"Software deployment; Computer science; Scalability; Big data; Distributed computing; Reliability (semiconductor); Fault tolerance; Process (computing); Pipeline (software); Data science; Systems engineering; Software engineering; Database; Engineering; Data mining","score_opus":0.5304000793599303,"score_gpt":0.36865175447630044,"score_spread":0.16174832488362983,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W4318147681","genre_codex":"methods","genre_gemma":"methods","domain_codex":null,"domain_gemma":null,"model_version":"metacan-v3-hybrid-931329e0061c","genre_candidate":"methods","genre_consensus":"methods","domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.0016123644,0.00025045883,0.9930333,0.00087429635,0.00006939592,0.00041247153,0.000059915063,0.00072437833,0.0029633997],"genre_scores_gemma":[0.05570433,0.00060885894,0.9397196,0.00025494088,0.00007829457,0.0010061642,0.00019722749,0.00024304436,0.0021875214],"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"theoretical_or_conceptual","domain_scores_codex":[0.9932532,0.0022594563,0.0006739167,0.00074746966,0.0025087327,0.000557156],"domain_scores_gemma":[0.99286026,0.0021383825,0.0007730268,0.0017285766,0.0016484444,0.00085128646],"candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.011489146,0.0016311746,0.0010150955,0.0017807591,0.0019706283,0.0067358688,0.005956908,0.0035606122,0.0025290584],"category_scores_gemma":[0.012633559,0.001466992,0.0019468664,0.0010093584,0.005299695,0.0055988543,0.005599977,0.0044317367,0.0012243085],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":"theoretical_or_conceptual","about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_system_candidate":false,"about_ca_system_consensus":false,"study_design_scores_codex":[0.000042586173,0.00014403158,0.0009847308,0.0005475085,0.000069675436,0.0005291897,0.0009571418,0.16115154,0.0056670234,0.7894374,0.004581101,0.035888165],"study_design_scores_gemma":[0.00009846102,0.00035997585,0.00043036492,0.0008004114,0.000095666604,0.00045014126,0.0006494134,0.5183591,0.0076939003,0.35036898,0.12056621,0.00012739433],"about_ca_topic_score_codex":0.005929203,"about_ca_topic_score_gemma":0.0057316483,"teacher_disagreement_score":0.011489146,"about_ca_system_score_codex":0.002779607,"about_ca_system_score_gemma":0.008199648,"threshold_uncertainty_score":0.060761154},"labels":[],"label_agreement":null},{"id":"W4318147794","doi":"10.1109/bigdata55660.2022.10020299","title":"Temporal Graph Representation Learning via Maximal Cliques","year":2022,"lang":"en","type":"article","venue":"2022 IEEE International Conference on Big Data (Big Data)","topic":"Advanced Graph Neural Networks","field":"Computer Science","cited_by":5,"is_retracted":false,"has_abstract":true,"route_ca_aff":true,"route_ca_fund":true,"route_ca_venue":false,"route_about_ca":false,"ca_institutions":"York University","funders":"Natural Sciences and Engineering Research Council of Canada","keywords":"Clique; Computer science; Graph; Node (physics); Theoretical computer science; Embedding; Representation (politics); Graph embedding; Feature learning; Artificial intelligence; Mathematics; Combinatorics","score_opus":0.304841879149363,"score_gpt":0.3731440942115506,"score_spread":0.06830221506218759,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W4318147794","genre_codex":"methods","genre_gemma":"empirical","domain_codex":null,"domain_gemma":null,"model_version":"metacan-v3-hybrid-931329e0061c","genre_candidate":"empirical","genre_consensus":null,"domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.0626722,0.0006402356,0.93019384,0.00057540456,0.000084001374,0.00012564761,0.0010989761,0.0015960718,0.003013665],"genre_scores_gemma":[0.67906034,0.000725839,0.30563715,0.0004025857,0.00017498311,0.00031633442,0.006678321,0.00041331112,0.00659115],"study_design_codex":"simulation_or_modeling","study_design_gemma":"theoretical_or_conceptual","domain_scores_codex":[0.9993224,0.00016938164,0.000028191478,0.00026952228,0.0001361195,0.000074305826],"domain_scores_gemma":[0.9982273,0.0008627985,0.00023611421,0.00030036553,0.00025883227,0.00011462294],"candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.00075723307,0.0009151363,0.0009099173,0.0021965422,0.0007691778,0.000856853,0.0017926542,0.0011560807,0.0031513043],"category_scores_gemma":[0.004555995,0.00054813793,0.0013120095,0.002474522,0.0007372856,0.0030860398,0.0012881536,0.0014677284,0.0006382394],"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":[0.0002769576,0.00030533344,0.007614385,0.00036102763,0.00027193822,0.00030026876,0.00038268953,0.5407816,0.0051632156,0.06339067,0.018522,0.36262992],"study_design_scores_gemma":[0.000010620049,0.000015456457,0.00034472303,0.000012423307,0.000015544514,0.000029552819,0.00002656566,0.96996605,0.00041701406,0.027835771,0.0013202094,0.0000060794873],"about_ca_topic_score_codex":0.009070338,"about_ca_topic_score_gemma":0.02029452,"teacher_disagreement_score":0.009070338,"about_ca_system_score_codex":0.0012730452,"about_ca_system_score_gemma":0.0009455015,"threshold_uncertainty_score":0.018035054},"labels":[],"label_agreement":null},{"id":"W4318147813","doi":"10.1109/bigdata55660.2022.10021073","title":"Systematic Analysis of Public Transit Data Availability in Canada","year":2022,"lang":"en","type":"article","venue":"2022 IEEE International Conference on Big Data (Big Data)","topic":"Traffic Prediction and Management Techniques","field":"Engineering","cited_by":0,"is_retracted":false,"has_abstract":true,"route_ca_aff":true,"route_ca_fund":false,"route_ca_venue":false,"route_about_ca":true,"ca_institutions":"Carleton University","funders":"","keywords":"Public transport; Computer science; Transit (satellite); Transport engineering; Data science; Engineering","score_opus":0.2616339857047169,"score_gpt":0.3015119091661445,"score_spread":0.0398779234614276,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W4318147813","genre_codex":"dataset","genre_gemma":"empirical","domain_codex":null,"domain_gemma":null,"model_version":"metacan-v3-hybrid-931329e0061c","genre_candidate":"empirical","genre_consensus":null,"domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.2736846,0.015515967,0.010051286,0.004478617,0.00034182257,0.0016983995,0.67497885,0.00073941104,0.018511027],"genre_scores_gemma":[0.6276519,0.008442282,0.01598907,0.0013220075,0.00006796268,0.0018167344,0.3389808,0.0003664286,0.0053629586],"study_design_codex":"observational","study_design_gemma":"observational","domain_scores_codex":[0.96091455,0.004096286,0.0040185773,0.0045242705,0.023429621,0.003016848],"domain_scores_gemma":[0.8352763,0.025023766,0.0134646,0.009590514,0.11312123,0.003523588],"candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.02141101,0.0009038795,0.0011761726,0.0196394,0.005032693,0.004196256,0.0030526796,0.0006851759,0.002207376],"category_scores_gemma":[0.10307331,0.0009335437,0.0013404449,0.054232754,0.0019962387,0.0013005597,0.0037525836,0.0012896936,0.00046515654],"study_design_candidate":"observational","study_design_consensus":"observational","about_ca_topic_candidate":true,"about_ca_topic_consensus":true,"about_ca_system_candidate":false,"about_ca_system_consensus":false,"study_design_scores_codex":[0.000383519,0.000095605355,0.7346997,0.0060779224,0.0015682151,0.0006033641,0.0073995334,0.003689566,0.0011407488,0.008620101,0.14995041,0.085771315],"study_design_scores_gemma":[0.000065534885,0.00004990445,0.79444987,0.0041393084,0.0007057446,0.00020631705,0.005833281,0.0054287007,0.0017237276,0.0009410449,0.18617232,0.00028424928],"about_ca_topic_score_codex":0.99421096,"about_ca_topic_score_gemma":0.995789,"teacher_disagreement_score":0.07039544,"about_ca_system_score_codex":0.07039544,"about_ca_system_score_gemma":0.21711516,"threshold_uncertainty_score":0.5107571},"labels":[],"label_agreement":null},{"id":"W4318147859","doi":"10.1109/bigdata55660.2022.10020507","title":"Discovering Limitations of Image Quality Assessments with Noised Deep Learning Image Sets","year":2022,"lang":"en","type":"article","venue":"2022 IEEE International Conference on Big Data (Big Data)","topic":"Image and Video Quality Assessment","field":"Computer Science","cited_by":4,"is_retracted":false,"has_abstract":true,"route_ca_aff":false,"route_ca_fund":true,"route_ca_venue":false,"route_about_ca":false,"ca_institutions":"","funders":"Canadian Institute for Advanced Research; Google","keywords":"MNIST database; Computer science; Artificial intelligence; Pixel; Image quality; Image (mathematics); Image processing; Computer vision; Noise (video); Image resolution; Image restoration; Pattern recognition (psychology); Deep learning","score_opus":0.4654309621175568,"score_gpt":0.43339241516499916,"score_spread":0.032038546952557645,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W4318147859","genre_codex":"methods","genre_gemma":"empirical","domain_codex":null,"domain_gemma":null,"model_version":"metacan-v3-hybrid-931329e0061c","genre_candidate":"empirical","genre_consensus":null,"domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.3446874,0.007715698,0.6301368,0.0029908875,0.00068517064,0.00049492624,0.0019400462,0.004990181,0.0063588372],"genre_scores_gemma":[0.74442685,0.0011509577,0.24855436,0.0010453566,0.0001491136,0.0002173892,0.0024018646,0.00041682043,0.0016373219],"study_design_codex":"design_other","study_design_gemma":"observational","domain_scores_codex":[0.99327326,0.0016991002,0.00060203334,0.0012704992,0.0028354174,0.0003197021],"domain_scores_gemma":[0.98134047,0.00810902,0.0015030673,0.003795022,0.004845674,0.00040690284],"candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.01108315,0.0012618053,0.0010315807,0.0020805527,0.0007885604,0.0024372311,0.0026810558,0.001659436,0.0009850285],"category_scores_gemma":[0.05345085,0.0005625257,0.00095652894,0.0011408797,0.0020807032,0.003599929,0.0025794352,0.0024134272,0.000592606],"study_design_candidate":"observational","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":[0.0022278386,0.00055257394,0.030076839,0.0015439971,0.0011026876,0.00042887582,0.00078244286,0.29053777,0.0635741,0.012187201,0.013184722,0.583801],"study_design_scores_gemma":[0.00008872569,0.0004946391,0.01073687,0.00024691195,0.00018755968,0.00044839445,0.00023269544,0.9001076,0.06316436,0.019689605,0.0045014583,0.00010120643],"about_ca_topic_score_codex":0.0061328434,"about_ca_topic_score_gemma":0.0059250332,"teacher_disagreement_score":0.01108315,"about_ca_system_score_codex":0.0018872306,"about_ca_system_score_gemma":0.0011095019,"threshold_uncertainty_score":0.058614016},"labels":[],"label_agreement":null},{"id":"W4318148079","doi":"10.1109/bigdata55660.2022.10020681","title":"Don’t Blindly Use Data: Towards a Data Statement for Computational Financial Research","year":2022,"lang":"en","type":"article","venue":"2022 IEEE International Conference on Big Data (Big Data)","topic":"Machine Learning and Data Classification","field":"Computer Science","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":"Dalhousie University","funders":"","keywords":"Computer science; Statement (logic); Data science; Data quality; Context (archaeology); Problem statement; Focus (optics); Quality (philosophy); Work (physics); Management science","score_opus":0.7268565196499841,"score_gpt":0.47980048120678986,"score_spread":0.24705603844319424,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W4318148079","genre_codex":"methods","genre_gemma":"methods","domain_codex":null,"domain_gemma":"reporting","model_version":"metacan-v3-hybrid-931329e0061c","genre_candidate":"methods","genre_consensus":"methods","domain_candidate":"reporting","domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.003764737,0.0032951653,0.80155927,0.17306444,0.0030135908,0.0014974774,0.0034347824,0.00094518834,0.009425396],"genre_scores_gemma":[0.06475701,0.003105255,0.8833661,0.032337423,0.003521182,0.0055062356,0.0047173095,0.0007203915,0.0019691244],"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"theoretical_or_conceptual","domain_scores_codex":[0.60454595,0.28253803,0.042431764,0.01332964,0.053953983,0.0032006348],"domain_scores_gemma":[0.18198788,0.6259692,0.03656094,0.07859792,0.07081633,0.0060677226],"candidate_categories":["metaresearch"],"consensus_categories":["metaresearch"],"category_scores_codex":[0.4443332,0.002405703,0.004070159,0.015944952,0.008088636,0.037685286,0.009397742,0.017163945,0.0058263973],"category_scores_gemma":[0.66883457,0.002919219,0.0032678624,0.02407745,0.033429917,0.07325903,0.021527793,0.032779187,0.0042723217],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":"theoretical_or_conceptual","about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_system_candidate":false,"about_ca_system_consensus":false,"study_design_scores_codex":[0.00022926608,0.0001200236,0.0031763853,0.0013379204,0.000076944874,0.00014772268,0.0039670523,0.0023781406,0.0003914054,0.86176986,0.055524774,0.07088062],"study_design_scores_gemma":[0.00010883845,0.000091252594,0.0005730687,0.0030850666,0.00003877255,0.00013756027,0.0023977503,0.009284213,0.0008814851,0.8529316,0.13032667,0.00014379491],"about_ca_topic_score_codex":0.0039968295,"about_ca_topic_score_gemma":0.0029608067,"teacher_disagreement_score":0.5556668,"about_ca_system_score_codex":0.008956787,"about_ca_system_score_gemma":0.03215597,"threshold_uncertainty_score":0.6852361},"labels":[],"label_agreement":null},{"id":"W4318148204","doi":"10.1109/bigdata55660.2022.10020871","title":"Predicting Deduplication Performance: An Analytical Model and Empirical Evaluation","year":2022,"lang":"en","type":"article","venue":"2022 IEEE International Conference on Big Data (Big Data)","topic":"Advanced Data Storage Technologies","field":"Computer Science","cited_by":3,"is_retracted":false,"has_abstract":true,"route_ca_aff":true,"route_ca_fund":false,"route_ca_venue":false,"route_about_ca":false,"ca_institutions":"University of Alberta","funders":"","keywords":"Data deduplication; Computer science; Data mining; Software versioning; Kernel (algebra); Database; Software; Operating system","score_opus":0.5054255254777182,"score_gpt":0.4297697773039301,"score_spread":0.07565574817378812,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W4318148204","genre_codex":"empirical","genre_gemma":"empirical","domain_codex":null,"domain_gemma":null,"model_version":"metacan-v3-hybrid-931329e0061c","genre_candidate":"empirical","genre_consensus":"empirical","domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.76583385,0.0045681745,0.2065862,0.0030623125,0.00022860475,0.0007240386,0.0054200767,0.008162227,0.0054144943],"genre_scores_gemma":[0.8925783,0.0016704658,0.09767785,0.0004612896,0.000097184755,0.0004259814,0.004873342,0.00036595107,0.0018497391],"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","domain_scores_codex":[0.99726653,0.000566824,0.00019250513,0.0007603342,0.00082037237,0.00039341557],"domain_scores_gemma":[0.9695179,0.02132299,0.0018756068,0.002342374,0.004431116,0.0005100341],"candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.005261728,0.0024371198,0.0013403287,0.0025385935,0.0008354508,0.0023080048,0.0025987716,0.0022693628,0.0012734107],"category_scores_gemma":[0.0354397,0.00096775516,0.0008685465,0.003184803,0.0011492601,0.0045909774,0.001050563,0.002639818,0.0012018873],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_system_candidate":false,"about_ca_system_consensus":false,"study_design_scores_codex":[0.000428867,0.0006129332,0.025926458,0.00039097114,0.0000989893,0.00016363653,0.00020066097,0.9233492,0.0036100706,0.002394123,0.0062687127,0.036555417],"study_design_scores_gemma":[0.0000081982425,0.000057173795,0.0011951282,0.000013400804,0.0000089169225,0.00004644111,0.000030312953,0.9963342,0.0012239654,0.00075557886,0.00031499375,0.00001167374],"about_ca_topic_score_codex":0.018316785,"about_ca_topic_score_gemma":0.012924536,"teacher_disagreement_score":0.018316785,"about_ca_system_score_codex":0.0034190482,"about_ca_system_score_gemma":0.0022904268,"threshold_uncertainty_score":0.036420345},"labels":[],"label_agreement":null},{"id":"W4318148225","doi":"10.1109/bigdata55660.2022.10020386","title":"Adaptive Method for Machine Learning Model Selection in Data Science Projects","year":2022,"lang":"en","type":"article","venue":"2022 IEEE International Conference on Big Data (Big Data)","topic":"Software Engineering Research","field":"Computer Science","cited_by":3,"is_retracted":false,"has_abstract":true,"route_ca_aff":true,"route_ca_fund":true,"route_ca_venue":false,"route_about_ca":false,"ca_institutions":"University of Waterloo","funders":"Natural Sciences and Engineering Research Council of Canada","keywords":"Computer science; Selection (genetic algorithm); Machine learning; Process (computing); Heuristics; Feature selection; Model selection; Artificial intelligence; Data mining","score_opus":0.4726800557815008,"score_gpt":0.4217361885031067,"score_spread":0.05094386727839412,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W4318148225","genre_codex":"methods","genre_gemma":"methods","domain_codex":null,"domain_gemma":null,"model_version":"metacan-v3-hybrid-931329e0061c","genre_candidate":"methods","genre_consensus":"methods","domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.0052037993,0.000044003194,0.99351317,0.00010168571,0.000013121596,0.00010563633,0.000041332525,0.0005492142,0.000428109],"genre_scores_gemma":[0.1561818,0.000062827356,0.8414322,0.0001569201,0.000033153538,0.000732776,0.00029894692,0.00020844261,0.00089298707],"study_design_codex":"simulation_or_modeling","study_design_gemma":"theoretical_or_conceptual","domain_scores_codex":[0.9905923,0.0060970583,0.0004585738,0.001169952,0.0014121787,0.00026998724],"domain_scores_gemma":[0.9591773,0.03396612,0.0015706968,0.0022822358,0.0026006137,0.0004029306],"candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.01356869,0.0010950305,0.00091748656,0.0029287112,0.0009752389,0.0018309184,0.0026372445,0.0016391269,0.0038153096],"category_scores_gemma":[0.048986614,0.0007623716,0.0017023725,0.0022778574,0.0009944342,0.0021379853,0.0023432525,0.0033485962,0.00066401967],"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":[0.00041866204,0.00037942908,0.012005607,0.00032979556,0.0004161185,0.00041006313,0.00082984666,0.5824137,0.0031299528,0.06175525,0.004448804,0.33346286],"study_design_scores_gemma":[0.000026063632,0.000030244668,0.00026473304,0.000016084405,0.00001652133,0.00002964246,0.00004130354,0.983095,0.00062661717,0.014736877,0.0011054692,0.000011431773],"about_ca_topic_score_codex":0.004615558,"about_ca_topic_score_gemma":0.0074260947,"teacher_disagreement_score":0.01356869,"about_ca_system_score_codex":0.0015683426,"about_ca_system_score_gemma":0.0028612197,"threshold_uncertainty_score":0.071758926},"labels":[],"label_agreement":null},{"id":"W4318148360","doi":"10.1109/bigdata55660.2022.10020874","title":"LSTM-based Pulmonary Air Leak Forecasting for Chest Tube Management","year":2022,"lang":"en","type":"article","venue":"2022 IEEE International Conference on Big Data (Big Data)","topic":"COVID-19 diagnosis using AI","field":"Medicine","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":"Ottawa Hospital","funders":"","keywords":"Context (archaeology); Empyema; Chest tube; Computer science; Leak; Autoregressive model; Medicine; Intensive care medicine; Surgery; Engineering","score_opus":0.44725369165867457,"score_gpt":0.3922616421471449,"score_spread":0.054992049511529695,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W4318148360","genre_codex":"methods","genre_gemma":"empirical","domain_codex":null,"domain_gemma":null,"model_version":"metacan-v3-hybrid-931329e0061c","genre_candidate":"empirical","genre_consensus":null,"domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.21465442,0.0038528657,0.76370835,0.0022124872,0.000783473,0.00012918419,0.0016694455,0.007079919,0.005909872],"genre_scores_gemma":[0.9515593,0.00056207,0.044656992,0.00026909247,0.00013021835,0.00007523856,0.0008037821,0.00004096017,0.00190225],"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","domain_scores_codex":[0.9998785,0.00002232141,0.00001235563,0.000033932793,0.000027321494,0.000025533822],"domain_scores_gemma":[0.99965787,0.00017049356,0.000034957277,0.000020558398,0.00009568693,0.000020495638],"candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0004170867,0.00071364536,0.0004885659,0.000470808,0.00021204988,0.0005162762,0.00074380456,0.00081891584,0.0014749538],"category_scores_gemma":[0.0015498409,0.00022867405,0.00041337562,0.0005207977,0.00015862366,0.0007426795,0.00039275776,0.0010080592,0.0005521683],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_system_candidate":false,"about_ca_system_consensus":false,"study_design_scores_codex":[0.000421488,0.0002964929,0.005473239,0.00017594233,0.00012171131,0.00027380526,0.000099030636,0.6764566,0.013546727,0.0014230192,0.0071902936,0.2945217],"study_design_scores_gemma":[0.000003431463,0.000019304458,0.0002717075,0.000006112888,0.000007870009,0.000012159284,0.000006173507,0.9979742,0.0008783244,0.0006159975,0.00020123247,0.0000035437406],"about_ca_topic_score_codex":0.006949858,"about_ca_topic_score_gemma":0.006893394,"teacher_disagreement_score":0.006949858,"about_ca_system_score_codex":0.0005592936,"about_ca_system_score_gemma":0.0007360314,"threshold_uncertainty_score":0.01381886},"labels":[],"label_agreement":null},{"id":"W4318148398","doi":"10.1109/bigdata55660.2022.10020567","title":"Heterogeneity in feature importance and prediction performance for sales at the market and store levels: the case of branded yogurt products in Quebec","year":2022,"lang":"en","type":"article","venue":"2022 IEEE International Conference on Big Data (Big Data)","topic":"Consumer Attitudes and Food Labeling","field":"Medicine","cited_by":2,"is_retracted":false,"has_abstract":true,"route_ca_aff":true,"route_ca_fund":false,"route_ca_venue":false,"route_about_ca":true,"ca_institutions":"Université de Montréal; McGill University","funders":"HORIZON EUROPE Health","keywords":"Gradient boosting; Random forest; Aggregate (composite); Python (programming language); Loyalty; Computer science; Loyalty business model; Product (mathematics); Aggregate data; Marketing; Machine learning; Business; Statistics","score_opus":0.248189087809298,"score_gpt":0.3494119663719226,"score_spread":0.10122287856262463,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W4318148398","genre_codex":"empirical","genre_gemma":"empirical","domain_codex":null,"domain_gemma":null,"model_version":"metacan-v3-hybrid-931329e0061c","genre_candidate":"empirical","genre_consensus":"empirical","domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.9753532,0.0009602467,0.002429241,0.0008159621,0.00006471211,0.00004485552,0.016336918,0.00040742412,0.0035873945],"genre_scores_gemma":[0.96634424,0.00015690563,0.0025686084,0.00013812372,0.000018019326,0.000018165205,0.027830908,0.000052198502,0.002872916],"study_design_codex":"observational","study_design_gemma":"observational","domain_scores_codex":[0.9993728,0.00009424137,0.00002881494,0.00020161156,0.00012179419,0.00018079086],"domain_scores_gemma":[0.9978549,0.0006293877,0.00011550462,0.00029333416,0.00090149563,0.00020527242],"candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0014526311,0.00070658873,0.00042658934,0.0010407636,0.0010674767,0.0012819073,0.0012182739,0.0005966685,0.0024415115],"category_scores_gemma":[0.004629441,0.00021113716,0.00075544347,0.0014968321,0.0005770521,0.00082854246,0.00058893225,0.0011887713,0.00082300836],"study_design_candidate":"observational","study_design_consensus":"observational","about_ca_topic_candidate":true,"about_ca_topic_consensus":true,"about_ca_system_candidate":false,"about_ca_system_consensus":false,"study_design_scores_codex":[0.0009962459,0.0004673126,0.80267954,0.00015331162,0.00047419252,0.00033059742,0.0004329208,0.05636028,0.0027260482,0.00078785594,0.04266669,0.091925114],"study_design_scores_gemma":[0.00007593433,0.00012196469,0.6335072,0.000055197426,0.0001409767,0.00012855676,0.0010271147,0.35097337,0.0016380643,0.0005573159,0.011691342,0.00008290732],"about_ca_topic_score_codex":0.9598614,"about_ca_topic_score_gemma":0.9728767,"teacher_disagreement_score":0.040138602,"about_ca_system_score_codex":0.0076895757,"about_ca_system_score_gemma":0.00485266,"threshold_uncertainty_score":0.08074993},"labels":[],"label_agreement":null},{"id":"W4318185035","doi":"10.1109/bigdata55660.2022.10020256","title":"Neuro-heuristic Pallet Detection for Automated Guided Vehicle Navigation","year":2022,"lang":"en","type":"article","venue":"2022 IEEE International Conference on Big Data (Big Data)","topic":"Advanced Manufacturing and Logistics Optimization","field":"Engineering","cited_by":11,"is_retracted":false,"has_abstract":true,"route_ca_aff":true,"route_ca_fund":false,"route_ca_venue":false,"route_about_ca":false,"ca_institutions":"Brandon University","funders":"Silesian University of Technology","keywords":"Heuristics; Computer science; Pallet; Focus (optics); Automation; Heuristic; Artificial intelligence; Computer vision; Point (geometry); Robot; Real-time computing; Engineering","score_opus":0.24122699971033065,"score_gpt":0.33973951328562285,"score_spread":0.0985125135752922,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W4318185035","genre_codex":"methods","genre_gemma":"empirical","domain_codex":null,"domain_gemma":null,"model_version":"metacan-v3-hybrid-931329e0061c","genre_candidate":"empirical","genre_consensus":null,"domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.0763492,0.0008982603,0.9126941,0.0004046042,0.0001528597,0.00017327421,0.00024716448,0.0015322833,0.007548164],"genre_scores_gemma":[0.8457364,0.00024680598,0.14830434,0.00023902782,0.00005823951,0.00014081744,0.00031869512,0.0000693331,0.004886358],"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","domain_scores_codex":[0.9996866,0.000065646615,0.000017456476,0.00009708948,0.00005998653,0.00007335118],"domain_scores_gemma":[0.9994672,0.000253171,0.00008186465,0.000028680686,0.00012651506,0.000042604046],"candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0005522354,0.000992346,0.00087968586,0.0011802482,0.00050683273,0.0011067679,0.00157526,0.0013896669,0.0017685571],"category_scores_gemma":[0.0013620808,0.0005541209,0.00060025573,0.00083411013,0.0006014222,0.00087820267,0.0006344344,0.00074786256,0.00037903627],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_system_candidate":false,"about_ca_system_consensus":false,"study_design_scores_codex":[0.00014904073,0.00010443282,0.0006794676,0.00004568798,0.00004382634,0.0000606751,0.000029531222,0.93812114,0.0010573576,0.001977157,0.0014013953,0.056330316],"study_design_scores_gemma":[0.000005603355,0.000018500483,0.00008515077,0.0000028983734,0.000004400482,0.000006773623,0.0000061494825,0.99883443,0.00026997292,0.00061400654,0.00014907587,0.0000030854053],"about_ca_topic_score_codex":0.019147463,"about_ca_topic_score_gemma":0.015391226,"teacher_disagreement_score":0.019147463,"about_ca_system_score_codex":0.001506178,"about_ca_system_score_gemma":0.0015648603,"threshold_uncertainty_score":0.03807199},"labels":[],"label_agreement":null},{"id":"W4318185064","doi":"10.1109/bigdata55660.2022.10020982","title":"Mixers Detection in bitcoin network: a step towards detecting money laundering in crypto-currencies","year":2022,"lang":"en","type":"article","venue":"2022 IEEE International Conference on Big Data (Big Data)","topic":"Crime, Illicit Activities, and Governance","field":"Social Sciences","cited_by":16,"is_retracted":false,"has_abstract":true,"route_ca_aff":true,"route_ca_fund":false,"route_ca_venue":false,"route_about_ca":false,"ca_institutions":"University of New Brunswick","funders":"Innovation Fund","keywords":"Money laundering; Computer science; Proof-of-work system; Anonymity; Database transaction; Computer security; Mixing (physics); Process (computing); Cryptocurrency; Communication source; Digital currency; Pruning; Tree (set theory); Decision tree; Computer network; Data mining; Business; World Wide Web; Operating system; Payment; Database","score_opus":0.3417629011562336,"score_gpt":0.3747394520079122,"score_spread":0.03297655085167861,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W4318185064","genre_codex":"methods","genre_gemma":"empirical","domain_codex":null,"domain_gemma":null,"model_version":"metacan-v3-hybrid-931329e0061c","genre_candidate":"empirical","genre_consensus":null,"domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.3775644,0.0010945365,0.6036886,0.0011290353,0.0002540206,0.0004053303,0.0012347437,0.008357229,0.006272116],"genre_scores_gemma":[0.9011925,0.0001887238,0.09359196,0.00017255392,0.00006294036,0.00011188042,0.0012089795,0.00007634619,0.0033940864],"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","domain_scores_codex":[0.99911016,0.00014422653,0.0000551813,0.00016655875,0.00034162862,0.00018230829],"domain_scores_gemma":[0.9988231,0.00034651134,0.0002238084,0.00011739304,0.0003765688,0.00011255178],"candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0010795204,0.0009716603,0.0009107617,0.002629901,0.0006217373,0.0013744634,0.00084614963,0.001122523,0.0015987204],"category_scores_gemma":[0.002326518,0.00020996876,0.00058200926,0.0011496986,0.00030344343,0.0020668043,0.0009095234,0.0013904248,0.0010235611],"study_design_candidate":"simulation_or_modeling","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":[0.0012236041,0.0011766376,0.06940215,0.00027386568,0.00018699194,0.0011727078,0.00032013195,0.09410695,0.052439637,0.008521674,0.01387938,0.7572962],"study_design_scores_gemma":[0.000029676878,0.00013752596,0.0067635514,0.00002888824,0.000033740584,0.00028655794,0.00011368763,0.96215767,0.022852002,0.0035969645,0.003964792,0.000034919056],"about_ca_topic_score_codex":0.0030404327,"about_ca_topic_score_gemma":0.0026937758,"teacher_disagreement_score":0.0030404327,"about_ca_system_score_codex":0.00059006555,"about_ca_system_score_gemma":0.00093218917,"threshold_uncertainty_score":0.0060454607},"labels":[],"label_agreement":null},{"id":"W4318185132","doi":"10.1109/bigdata55660.2022.10020719","title":"From Data Warehouse to Lakehouse: A Comparative Review","year":2022,"lang":"en","type":"review","venue":"2022 IEEE International Conference on Big Data (Big Data)","topic":"Data Quality and Management","field":"Decision Sciences","cited_by":65,"is_retracted":false,"has_abstract":true,"route_ca_aff":true,"route_ca_fund":false,"route_ca_venue":false,"route_about_ca":false,"ca_institutions":"Queen's University","funders":"","keywords":"Data warehouse; Computer science; Big data; Data science; Unstructured data; Data management; Data extraction; Analytics; Strengths and weaknesses; Data transformation; Data virtualization; Database; Data mining; Cloud computing","score_opus":0.9445247924110595,"score_gpt":0.5941263284283097,"score_spread":0.35039846398274976,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W4318185132","genre_codex":"review","genre_gemma":"review","domain_codex":null,"domain_gemma":null,"model_version":"metacan-v3-hybrid-931329e0061c","genre_candidate":"review","genre_consensus":"review","domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.00010721733,0.9978928,0.00030703153,0.00048495387,0.00010704181,0.000010201019,0.000029078241,0.000008015,0.0010536273],"genre_scores_gemma":[0.0007362419,0.99817336,0.000569218,0.00025261138,0.000090505775,0.000012099252,0.000043313183,0.0000040381738,0.0001185109],"study_design_codex":"design_other","study_design_gemma":"not_applicable","domain_scores_codex":[0.9984487,0.0003991753,0.00029800634,0.00020314187,0.00056065846,0.000090239984],"domain_scores_gemma":[0.9920392,0.005211985,0.0005698972,0.00016442945,0.0017865688,0.00022792224],"candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.004710013,0.0008632672,0.0016209617,0.008820729,0.00055671204,0.0029137474,0.0021023443,0.0013763786,0.003942612],"category_scores_gemma":[0.007918675,0.0007103677,0.001130129,0.01868758,0.001201794,0.005366818,0.0014910166,0.0017421477,0.0016465914],"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":[0.00009266986,0.000077797624,0.00053986185,0.04937923,0.00017405619,0.00011664714,0.00028627395,0.00041383723,0.0003335404,0.014120505,0.03354496,0.9009205],"study_design_scores_gemma":[0.000017727021,0.00011718158,0.0012627111,0.024655776,0.00025683423,0.0005879574,0.00041804864,0.00017196625,0.0003118061,0.004335397,0.96782255,0.000042080053],"about_ca_topic_score_codex":0.0041878875,"about_ca_topic_score_gemma":0.0065397513,"teacher_disagreement_score":0.008820729,"about_ca_system_score_codex":0.0018406397,"about_ca_system_score_gemma":0.004729199,"threshold_uncertainty_score":0.024909258},"labels":[],"label_agreement":null},{"id":"W4318185157","doi":"10.1109/bigdata55660.2022.10020564","title":"A Novel Rigorous Measurement Model for Big Data Quality Characteristics","year":2022,"lang":"en","type":"article","venue":"2022 IEEE International Conference on Big Data (Big Data)","topic":"Big Data and Business Intelligence","field":"Business, Management and Accounting","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":"Concordia University","funders":"","keywords":"Big data; Computer science; Data modeling; Quality (philosophy); Data mining; Database; Physics","score_opus":0.7505477363862263,"score_gpt":0.4080711581157322,"score_spread":0.3424765782704941,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W4318185157","genre_codex":"methods","genre_gemma":"methods","domain_codex":null,"domain_gemma":null,"model_version":"metacan-v3-hybrid-931329e0061c","genre_candidate":"methods","genre_consensus":"methods","domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.0042180596,0.00007427587,0.99337924,0.00031405236,0.000020638063,0.00012611625,0.00010694941,0.00033067507,0.0014299435],"genre_scores_gemma":[0.41778207,0.000395852,0.57740724,0.0002587934,0.00012345242,0.0011441195,0.0007260084,0.00012785438,0.002034652],"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"theoretical_or_conceptual","domain_scores_codex":[0.9797633,0.0062470683,0.0019002126,0.004127363,0.00700897,0.00095312344],"domain_scores_gemma":[0.9761101,0.009616345,0.0031940846,0.002748327,0.0077815107,0.00054961117],"candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.012654334,0.0015311678,0.001163578,0.0047482722,0.0010908226,0.0057866126,0.0036667243,0.0018103351,0.0020580892],"category_scores_gemma":[0.03952235,0.00080433075,0.002152782,0.0044514914,0.0027171138,0.012623582,0.0035074262,0.0032030542,0.0006707771],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":"theoretical_or_conceptual","about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_system_candidate":false,"about_ca_system_consensus":false,"study_design_scores_codex":[0.0001348605,0.00027713118,0.015199265,0.00037093306,0.0002345721,0.00024564267,0.0010130688,0.2595353,0.0048446083,0.59527856,0.004046548,0.11881947],"study_design_scores_gemma":[0.000017064209,0.00007747099,0.0018590015,0.000034539782,0.000049663347,0.00008547423,0.000101002675,0.89379555,0.0011323751,0.10052442,0.0022810625,0.00004238749],"about_ca_topic_score_codex":0.01017129,"about_ca_topic_score_gemma":0.005695426,"teacher_disagreement_score":0.012654334,"about_ca_system_score_codex":0.004878956,"about_ca_system_score_gemma":0.005008986,"threshold_uncertainty_score":0.06692332},"labels":[],"label_agreement":null},{"id":"W4318185484","doi":"10.1109/bigdata55660.2022.10020371","title":"Understanding Levels of Automation in Human-Machine Collaboration","year":2022,"lang":"en","type":"article","venue":"2022 IEEE International Conference on Big Data (Big Data)","topic":"Human-Automation Interaction and Safety","field":"Psychology","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":"University of Waterloo","funders":"","keywords":"Automation; Variety (cybernetics); Computer science; Autonomy; Process (computing); Task (project management); Data science; Risk analysis (engineering); Knowledge management; Systems engineering; Artificial intelligence; Engineering","score_opus":0.6250192258128631,"score_gpt":0.46840426658816997,"score_spread":0.1566149592246931,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W4318185484","genre_codex":"empirical","genre_gemma":"empirical","domain_codex":null,"domain_gemma":null,"model_version":"metacan-v3-hybrid-931329e0061c","genre_candidate":"empirical","genre_consensus":"empirical","domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.67729074,0.0036968938,0.20501277,0.005572604,0.00010832954,0.0002940454,0.00013690653,0.00022569358,0.107661955],"genre_scores_gemma":[0.98933053,0.0003222475,0.009654126,0.00007521779,0.00001944669,0.000056174464,0.000029573137,0.000012975111,0.0004997454],"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"observational","domain_scores_codex":[0.99152124,0.0044800085,0.0004620732,0.000998809,0.0017969081,0.000740978],"domain_scores_gemma":[0.96726036,0.019879274,0.0057195034,0.002125176,0.002951152,0.0020645128],"candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.006928116,0.00056443043,0.0003119305,0.0030368303,0.002092004,0.006474771,0.00081611646,0.0011681446,0.0026759848],"category_scores_gemma":[0.024886621,0.0005445731,0.0005362457,0.0012376298,0.00708887,0.008912162,0.0042715487,0.0013455073,0.00044937883],"study_design_candidate":"observational","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":[0.00037848385,0.00060459605,0.2148357,0.0012033029,0.00038410843,0.00045376195,0.11327646,0.026591606,0.00748954,0.45577985,0.0027351757,0.17626742],"study_design_scores_gemma":[0.00006768142,0.00046924697,0.21315001,0.0005744522,0.00013889221,0.00045950635,0.041337352,0.03164074,0.002238176,0.6760964,0.033628333,0.00019926265],"about_ca_topic_score_codex":0.0022560745,"about_ca_topic_score_gemma":0.0015540725,"teacher_disagreement_score":0.006928116,"about_ca_system_score_codex":0.0018125889,"about_ca_system_score_gemma":0.0017934453,"threshold_uncertainty_score":0.03663981},"labels":[],"label_agreement":null},{"id":"W4318186065","doi":"10.1109/bigdata55660.2022.10020878","title":"Cough Classification with Deep Derived Features using Audio Spectrogram Transformer","year":2022,"lang":"en","type":"article","venue":"2022 IEEE International Conference on Big Data (Big Data)","topic":"Respiratory and Cough-Related Research","field":"Medicine","cited_by":5,"is_retracted":false,"has_abstract":true,"route_ca_aff":true,"route_ca_fund":false,"route_ca_venue":false,"route_about_ca":false,"ca_institutions":"University of Ottawa; Carleton University; National Research Council Canada","funders":"National Research Council","keywords":"Spectrogram; Transformer; Computer science; Artificial intelligence; Feature engineering; Feature selection; Machine learning; Nonlinear system; Feature extraction; Deep learning; Data mining; Pattern recognition (psychology); Engineering","score_opus":0.44863126012549726,"score_gpt":0.407795632811187,"score_spread":0.04083562731431023,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W4318186065","genre_codex":"methods","genre_gemma":"empirical","domain_codex":null,"domain_gemma":null,"model_version":"metacan-v3-hybrid-931329e0061c","genre_candidate":"empirical","genre_consensus":null,"domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.27148947,0.0007474406,0.72180766,0.00039837524,0.00016542504,0.00008846479,0.00059884734,0.0026401426,0.0020642262],"genre_scores_gemma":[0.9548069,0.0002351524,0.04242563,0.00007819542,0.000041563744,0.000046990222,0.000684419,0.000043925695,0.001637203],"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","domain_scores_codex":[0.9998729,0.000027492435,0.000009788362,0.00003281273,0.000031960713,0.00002506196],"domain_scores_gemma":[0.9996983,0.00016185096,0.000025349083,0.000030479823,0.00006605963,0.000018031575],"candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.00043963984,0.00070511556,0.0004146312,0.0006107163,0.00013914549,0.0006013428,0.00048588312,0.00045363588,0.0010028023],"category_scores_gemma":[0.0013393456,0.00016252892,0.0006497673,0.00036963864,0.00020485488,0.00072150264,0.00062482746,0.000698486,0.00046907377],"study_design_candidate":"simulation_or_modeling","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":[0.00093537243,0.00046773526,0.01539025,0.00018434503,0.00018053122,0.00047181844,0.00012720493,0.2976134,0.06998735,0.0019191286,0.0034789215,0.6092439],"study_design_scores_gemma":[0.000007905277,0.000078713376,0.0016555262,0.0000066877274,0.000017649869,0.00005936857,0.00002024662,0.9904171,0.0065840604,0.0007731886,0.00037102558,0.00000857174],"about_ca_topic_score_codex":0.0023177422,"about_ca_topic_score_gemma":0.0024073746,"teacher_disagreement_score":0.0023177422,"about_ca_system_score_codex":0.00031800682,"about_ca_system_score_gemma":0.0003913396,"threshold_uncertainty_score":0.0046084523},"labels":[],"label_agreement":null},{"id":"W4318186088","doi":"10.1109/bigdata55660.2022.10020426","title":"A Convergence Theory for Federated Average: Beyond Smoothness","year":2022,"lang":"en","type":"article","venue":"2022 IEEE International Conference on Big Data (Big Data)","topic":"Privacy-Preserving Technologies in Data","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":"McGill University; University of British Columbia Hospital","funders":"","keywords":"Smoothness; Computer science; Lipschitz continuity; Convergence (economics); Bounded function; Function (biology); Gradient descent; Mathematical optimization; Scheme (mathematics); Stochastic gradient descent; Applied mathematics; Mathematics; Artificial intelligence; Artificial neural network; Mathematical analysis","score_opus":0.27825448653320733,"score_gpt":0.35604696893970694,"score_spread":0.07779248240649961,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W4318186088","genre_codex":"methods","genre_gemma":"empirical","domain_codex":null,"domain_gemma":null,"model_version":"metacan-v3-hybrid-931329e0061c","genre_candidate":"empirical","genre_consensus":null,"domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.006333806,0.00085195526,0.98880553,0.0008049782,0.000091174974,0.00006134855,0.00011686963,0.00028718018,0.002647219],"genre_scores_gemma":[0.57513344,0.0052369433,0.40154812,0.0018331729,0.00075009133,0.0011638778,0.00088226394,0.0011375998,0.012314585],"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"theoretical_or_conceptual","domain_scores_codex":[0.99538285,0.0015749136,0.0002561368,0.0010741407,0.0012694554,0.00044246108],"domain_scores_gemma":[0.9679183,0.022436444,0.0015400148,0.0025392561,0.004609753,0.0009562347],"candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.012870615,0.0029721358,0.0033089328,0.0031864007,0.0018129271,0.0031985678,0.00425962,0.0029644237,0.0053844503],"category_scores_gemma":[0.054118343,0.0010687755,0.0038356527,0.0029387341,0.0048205056,0.007911771,0.005842644,0.008657575,0.0012136098],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":"theoretical_or_conceptual","about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_system_candidate":false,"about_ca_system_consensus":false,"study_design_scores_codex":[0.00033432184,0.00014509141,0.0023045524,0.00052302214,0.00018555764,0.00027211194,0.0003971181,0.39244592,0.0022485629,0.5508825,0.0070860726,0.043175153],"study_design_scores_gemma":[0.000017207682,0.00006944041,0.00020721943,0.000073232695,0.00003350392,0.00008327382,0.000038043305,0.86092764,0.00073486404,0.13624458,0.0015457389,0.000025321],"about_ca_topic_score_codex":0.003876083,"about_ca_topic_score_gemma":0.0020290185,"teacher_disagreement_score":0.012870615,"about_ca_system_score_codex":0.003395754,"about_ca_system_score_gemma":0.0031315952,"threshold_uncertainty_score":0.06806713},"labels":[],"label_agreement":null},{"id":"W4318186095","doi":"10.1109/bigdata55660.2022.10020676","title":"DelphAI: A human-centered approach to time-series forecasting","year":2022,"lang":"en","type":"article","venue":"2022 IEEE International Conference on Big Data (Big Data)","topic":"Forecasting Techniques and Applications","field":"Decision Sciences","cited_by":1,"is_retracted":false,"has_abstract":true,"route_ca_aff":true,"route_ca_fund":false,"route_ca_venue":false,"route_about_ca":false,"ca_institutions":"University of Guelph; Vector Institute","funders":"","keywords":"Series (stratigraphy); Time series; Computer science; Machine learning; Geology","score_opus":0.762746653460341,"score_gpt":0.45772942114912724,"score_spread":0.30501723231121375,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W4318186095","genre_codex":"methods","genre_gemma":"empirical","domain_codex":null,"domain_gemma":null,"model_version":"metacan-v3-hybrid-931329e0061c","genre_candidate":"empirical","genre_consensus":null,"domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.0009555132,0.00005835968,0.9923248,0.0010571671,0.000044519307,0.00022544758,0.00010340294,0.000630876,0.004599962],"genre_scores_gemma":[0.06039764,0.00019592763,0.9342922,0.00050497375,0.00009578666,0.0013765328,0.00029086476,0.00036544105,0.0024805432],"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"simulation_or_modeling","domain_scores_codex":[0.97329235,0.019785771,0.00083618454,0.002463599,0.0031993517,0.00042267804],"domain_scores_gemma":[0.96563363,0.02360318,0.0019330285,0.004437908,0.003509086,0.00088310544],"candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.030983629,0.001733542,0.0010786287,0.0026197808,0.0020576094,0.0074803717,0.005349159,0.0028166345,0.00880087],"category_scores_gemma":[0.049414363,0.001253461,0.001871035,0.001989471,0.004426493,0.0050405595,0.008106864,0.005500954,0.0033617876],"study_design_candidate":"simulation_or_modeling","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":[0.0002741196,0.00056417653,0.0036832043,0.0011088607,0.00025782202,0.0005738488,0.0102959145,0.08584151,0.005907151,0.64379615,0.013861195,0.23383605],"study_design_scores_gemma":[0.00008399101,0.00021772005,0.00075452763,0.00038384146,0.00007586538,0.00024243601,0.0016253203,0.3706576,0.005063135,0.5429016,0.077858694,0.0001353933],"about_ca_topic_score_codex":0.002070817,"about_ca_topic_score_gemma":0.0025403448,"teacher_disagreement_score":0.030983629,"about_ca_system_score_codex":0.00251004,"about_ca_system_score_gemma":0.0052663474,"threshold_uncertainty_score":0.16385901},"labels":[],"label_agreement":null},{"id":"W4318187137","doi":"10.1109/bigdata55660.2022.10020293","title":"Entity Matching with AUC-Based Fairness","year":2022,"lang":"en","type":"article","venue":"2022 IEEE International Conference on Big Data (Big Data)","topic":"Data Quality and Management","field":"Decision Sciences","cited_by":6,"is_retracted":false,"has_abstract":true,"route_ca_aff":true,"route_ca_fund":false,"route_ca_venue":false,"route_about_ca":false,"ca_institutions":"Western University","funders":"","keywords":"Computer science; Matching (statistics); Statistics; Mathematics","score_opus":0.6664563814827509,"score_gpt":0.4557380255016488,"score_spread":0.21071835598110206,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W4318187137","genre_codex":"methods","genre_gemma":"empirical","domain_codex":null,"domain_gemma":null,"model_version":"metacan-v3-hybrid-931329e0061c","genre_candidate":"empirical","genre_consensus":null,"domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.037132047,0.00081251364,0.95734555,0.00075394026,0.000096234144,0.00014863956,0.00026444523,0.0010550558,0.0023916317],"genre_scores_gemma":[0.70722693,0.0003566648,0.2873195,0.00051867403,0.00020464674,0.00022845065,0.0007127209,0.00037595312,0.003056445],"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","domain_scores_codex":[0.9803161,0.009162597,0.0012787895,0.005053448,0.0032304153,0.0009585649],"domain_scores_gemma":[0.93071944,0.04092634,0.004897801,0.016207794,0.005929904,0.001318798],"candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.026785351,0.0012453592,0.002721226,0.0031801276,0.002043812,0.0050036134,0.0039262758,0.0031943105,0.002434625],"category_scores_gemma":[0.13645114,0.0006941913,0.0014895217,0.004596928,0.003615809,0.008819884,0.0066427207,0.0030064373,0.0008215393],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_system_candidate":false,"about_ca_system_consensus":false,"study_design_scores_codex":[0.00089988665,0.00024368936,0.021934552,0.00023818508,0.0002861011,0.0002996953,0.000722184,0.57170427,0.0023005242,0.13289781,0.0054176617,0.2630554],"study_design_scores_gemma":[0.00002857802,0.00008540067,0.0013470734,0.000029211482,0.000032515436,0.00016792584,0.00006326329,0.8687935,0.0031717373,0.12376944,0.002472335,0.00003892709],"about_ca_topic_score_codex":0.0037303858,"about_ca_topic_score_gemma":0.0020426258,"teacher_disagreement_score":0.026785351,"about_ca_system_score_codex":0.0030007774,"about_ca_system_score_gemma":0.0032368263,"threshold_uncertainty_score":0.14165622},"labels":[],"label_agreement":null},{"id":"W4318187617","doi":"10.1109/bigdata55660.2022.10020838","title":"A Semisupervised Approach to Predicting a Twitch Streamer’s Growth based on the Game Streamed","year":2022,"lang":"en","type":"article","venue":"2022 IEEE International Conference on Big Data (Big Data)","topic":"Consumer Market Behavior and Pricing","field":"Business, Management and Accounting","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":"Carleton University","funders":"","keywords":"Entertainment; Computer science; Meaning (existential); Set (abstract data type); Order (exchange); Approximation error; Work (physics); Artificial intelligence; Simulation; Algorithm; Economics; Engineering","score_opus":0.29481025468805927,"score_gpt":0.3133429269993256,"score_spread":0.01853267231126632,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W4318187617","genre_codex":"methods","genre_gemma":"empirical","domain_codex":null,"domain_gemma":null,"model_version":"metacan-v3-hybrid-931329e0061c","genre_candidate":"empirical","genre_consensus":null,"domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.40051043,0.00044305873,0.5903531,0.0014391495,0.000101957645,0.00026568328,0.0011328039,0.0014730188,0.0042809057],"genre_scores_gemma":[0.92024326,0.000109341716,0.075267926,0.00014606764,0.00007375338,0.0001254207,0.0010932765,0.00006205679,0.0028790226],"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","domain_scores_codex":[0.9992132,0.00030060657,0.000050058203,0.0002462058,0.00011455181,0.00007533125],"domain_scores_gemma":[0.99540174,0.0027415084,0.0006631749,0.00022772585,0.0007574411,0.00020836643],"candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0016372046,0.00074632163,0.0007132775,0.0014695002,0.00035354946,0.0010018913,0.0014440983,0.00093818846,0.0011274017],"category_scores_gemma":[0.0064336997,0.00038097563,0.00055271434,0.0007737343,0.0004758436,0.0012726664,0.0005309081,0.0010714603,0.0005200967],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_system_candidate":false,"about_ca_system_consensus":false,"study_design_scores_codex":[0.00044917254,0.000760645,0.054732285,0.00012869832,0.00022277533,0.0002613897,0.00029258113,0.77768815,0.0034521827,0.0059619476,0.0061805397,0.14986971],"study_design_scores_gemma":[0.0000022139661,0.000014649405,0.0009938309,0.000002395339,0.0000031814964,0.000011861967,0.000010751639,0.9978169,0.00017794251,0.00084100314,0.00012198929,0.0000033604995],"about_ca_topic_score_codex":0.009389636,"about_ca_topic_score_gemma":0.0141084585,"teacher_disagreement_score":0.009389636,"about_ca_system_score_codex":0.001062446,"about_ca_system_score_gemma":0.00077650166,"threshold_uncertainty_score":0.018669963},"labels":[],"label_agreement":null},{"id":"W4318187680","doi":"10.1109/bigdata55660.2022.10020691","title":"Air Pollution Surveillance System: A Big Data Approach to Monitoring Adverse Health Outcomes for Public Health Interventions","year":2022,"lang":"en","type":"article","venue":"2022 IEEE International Conference on Big Data (Big Data)","topic":"Air Quality and Health Impacts","field":"Environmental Science","cited_by":3,"is_retracted":false,"has_abstract":true,"route_ca_aff":true,"route_ca_fund":true,"route_ca_venue":false,"route_about_ca":false,"ca_institutions":"University of Waterloo","funders":"Natural Sciences and Engineering Research Council of Canada","keywords":"Big data; Harm; Air pollution; Public health; Psychological intervention; Environmental health; Environmental planning; Business; Risk analysis (engineering); Computer science; Environmental science; Medicine; Political science; Data mining","score_opus":0.6416240419164759,"score_gpt":0.4430536513876135,"score_spread":0.19857039052886244,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W4318187680","genre_codex":"methods","genre_gemma":"empirical","domain_codex":null,"domain_gemma":null,"model_version":"metacan-v3-hybrid-931329e0061c","genre_candidate":"empirical","genre_consensus":null,"domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.09121646,0.0050360183,0.7058788,0.02353204,0.0024628928,0.0052873367,0.085301735,0.044299744,0.036984984],"genre_scores_gemma":[0.6280248,0.0028435353,0.3125728,0.005617256,0.0011513315,0.0022829715,0.03887006,0.00039392733,0.008243287],"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","domain_scores_codex":[0.99820995,0.00050487113,0.00020073273,0.00047478356,0.0004983586,0.000111268746],"domain_scores_gemma":[0.99701333,0.00088470796,0.00034652336,0.0005087175,0.0009107008,0.00033602695],"candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.003079797,0.0014416203,0.0014009925,0.0033672317,0.0010390051,0.0021800906,0.0023189378,0.0015899158,0.0025074815],"category_scores_gemma":[0.004894571,0.00047636215,0.0010529978,0.0021067827,0.0005059511,0.0023497269,0.00250697,0.0017563242,0.0010506585],"study_design_candidate":"simulation_or_modeling","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":[0.0014283697,0.0025651448,0.162631,0.0019410338,0.002172753,0.0007890032,0.0007986997,0.066451,0.016886167,0.028030368,0.19627991,0.52002656],"study_design_scores_gemma":[0.0002590033,0.0007958827,0.051084135,0.0003460101,0.00065047015,0.00038265763,0.00065371173,0.7879407,0.01628886,0.04663721,0.09478208,0.00017936034],"about_ca_topic_score_codex":0.011975295,"about_ca_topic_score_gemma":0.015757835,"teacher_disagreement_score":0.011975295,"about_ca_system_score_codex":0.0011001754,"about_ca_system_score_gemma":0.0024068845,"threshold_uncertainty_score":0.023811162},"labels":[],"label_agreement":null},{"id":"W4320024021","doi":"10.1109/bigdata55660.2022.10020313","title":"The Analysis and Development of an XAI Process on Feature Contribution Explanation","year":2022,"lang":"en","type":"article","venue":"2022 IEEE International Conference on Big Data (Big Data)","topic":"Explainable Artificial Intelligence (XAI)","field":"Computer Science","cited_by":25,"is_retracted":false,"has_abstract":true,"route_ca_aff":true,"route_ca_fund":false,"route_ca_venue":false,"route_about_ca":false,"ca_institutions":"Concordia University","funders":"","keywords":"Consistency (knowledge bases); Computer science; Feature (linguistics); Process (computing); Ranking (information retrieval); Data mining; Feature selection; Artificial intelligence; Information retrieval","score_opus":0.2539979832914212,"score_gpt":0.37115651489714635,"score_spread":0.11715853160572515,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W4320024021","genre_codex":"methods","genre_gemma":"methods","domain_codex":null,"domain_gemma":null,"model_version":"metacan-v3-hybrid-931329e0061c","genre_candidate":"methods","genre_consensus":"methods","domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.008737051,0.00016841065,0.98569226,0.0008166755,0.00002006271,0.00093720475,0.00019733225,0.0012508869,0.0021801493],"genre_scores_gemma":[0.048423726,0.00010965604,0.9495083,0.00008739282,0.000017054615,0.0006830896,0.00036965,0.00021396272,0.00058706035],"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","domain_scores_codex":[0.96217287,0.021078149,0.0026176213,0.0034295972,0.0099439155,0.0007579183],"domain_scores_gemma":[0.7786217,0.1443537,0.010000917,0.030777559,0.034967754,0.0012784117],"candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.057389364,0.0015796,0.0012249238,0.008872311,0.0024795318,0.0063423915,0.0037442874,0.0019257172,0.0037333197],"category_scores_gemma":[0.16814357,0.0014805228,0.002627977,0.005268311,0.003590911,0.010094336,0.006267645,0.0048918193,0.0010967541],"study_design_candidate":"simulation_or_modeling","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":[0.00021767002,0.000527631,0.019339943,0.0014574717,0.00024026439,0.00039301446,0.015920315,0.028682781,0.012136871,0.25358623,0.0051087523,0.66238904],"study_design_scores_gemma":[0.00014101448,0.00072500476,0.0136100035,0.0016301008,0.0002804664,0.00092558516,0.0055667087,0.5588713,0.05667326,0.2873986,0.07385308,0.00032477977],"about_ca_topic_score_codex":0.005007095,"about_ca_topic_score_gemma":0.0038783725,"teacher_disagreement_score":0.057389364,"about_ca_system_score_codex":0.0044024508,"about_ca_system_score_gemma":0.008682775,"threshold_uncertainty_score":0.30350757},"labels":[],"label_agreement":null},{"id":"W4320024064","doi":"10.1109/bigdata55660.2022.10020783","title":"Imbalanced Multi-layer Cloud Classification with Advanced Baseline Imager (ABI) and CloudSat/CALIPSO Data","year":2022,"lang":"en","type":"article","venue":"2022 IEEE International Conference on Big Data (Big Data)","topic":"Remote Sensing in Agriculture","field":"Environmental 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":"National Research Council Canada","funders":"National Security Agency; National Science Foundation","keywords":"Cloud computing; Random forest; Lidar; Computer science; Classifier (UML); Perceptron; Baseline (sea); Deep learning; Resampling; Remote sensing; Cloud cover; Cloud fraction; Artificial intelligence; Environmental science; Artificial neural network; Geography; Geology","score_opus":0.2510556575753,"score_gpt":0.3345477456942415,"score_spread":0.08349208811894149,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W4320024064","genre_codex":"empirical","genre_gemma":"empirical","domain_codex":null,"domain_gemma":null,"model_version":"metacan-v3-hybrid-931329e0061c","genre_candidate":"empirical","genre_consensus":"empirical","domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.89879465,0.0004287933,0.091090254,0.0005656497,0.00023136937,0.00022307581,0.0030093417,0.0029942465,0.0026626617],"genre_scores_gemma":[0.94267017,0.000087378496,0.05162381,0.00011761841,0.000054129527,0.00009422331,0.0044870675,0.000055426648,0.0008100708],"study_design_codex":"simulation_or_modeling","study_design_gemma":"observational","domain_scores_codex":[0.99948007,0.000068698944,0.00003152246,0.00012747702,0.0001654531,0.00012666194],"domain_scores_gemma":[0.9993895,0.00011920367,0.00009346495,0.00014478648,0.00019187314,0.000060998595],"candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0011073279,0.0009873868,0.0005153013,0.0009360527,0.0005159648,0.0007004881,0.0010647415,0.0006474014,0.0008789645],"category_scores_gemma":[0.0019027286,0.00022547932,0.00092678866,0.0010255553,0.0003295546,0.0013816733,0.0006656392,0.0009923138,0.000382585],"study_design_candidate":"observational","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":[0.0014520191,0.0019324062,0.09870572,0.00018908553,0.00037999515,0.0004675797,0.00021147565,0.58553976,0.0322347,0.0019928985,0.013214773,0.2636795],"study_design_scores_gemma":[0.00002036485,0.000060494396,0.011073071,0.0000044334834,0.000015352416,0.000024963152,0.000034890913,0.9829153,0.0047899233,0.0004645139,0.0005847741,0.0000118317475],"about_ca_topic_score_codex":0.020997876,"about_ca_topic_score_gemma":0.018551638,"teacher_disagreement_score":0.020997876,"about_ca_system_score_codex":0.0008138054,"about_ca_system_score_gemma":0.0007845617,"threshold_uncertainty_score":0.041751325},"labels":[],"label_agreement":null},{"id":"W4320024084","doi":"10.1109/bigdata55660.2022.10020739","title":"The Dimensional Analysis of Data Flow Programs That Include Multidimensional and User-Defined Functions","year":2022,"lang":"en","type":"article","venue":"2022 IEEE International Conference on Big Data (Big Data)","topic":"Scientific Computing and Data Management","field":"Decision Sciences","cited_by":0,"is_retracted":false,"has_abstract":true,"route_ca_aff":true,"route_ca_fund":true,"route_ca_venue":false,"route_about_ca":false,"ca_institutions":"University of Victoria","funders":"University of Victoria","keywords":"Computer science; Dimension (graph theory); Curse of dimensionality; Multidimensional analysis; Data mining; Flow (mathematics); Rank (graph theory); Theoretical computer science; Data flow diagram; Raw data; Artificial intelligence; Database; Programming language; Mathematics; Statistics","score_opus":0.6122670306410279,"score_gpt":0.4310872843677545,"score_spread":0.18117974627327343,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W4320024084","genre_codex":"methods","genre_gemma":"methods","domain_codex":null,"domain_gemma":null,"model_version":"metacan-v3-hybrid-931329e0061c","genre_candidate":"methods","genre_consensus":"methods","domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.0020252473,0.00016917498,0.99536175,0.00033619412,0.000032364558,0.00009266515,0.000096932185,0.0009850571,0.0009006074],"genre_scores_gemma":[0.03765425,0.0002877321,0.9587665,0.00034846395,0.00010141725,0.0004325172,0.00038672806,0.0006483166,0.0013740127],"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"simulation_or_modeling","domain_scores_codex":[0.9879171,0.004846397,0.001342913,0.0017226489,0.0035201823,0.00065071386],"domain_scores_gemma":[0.972822,0.017119665,0.0019072256,0.004471659,0.003319425,0.0003601137],"candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.013667611,0.0020047207,0.0015175034,0.0033737414,0.0018847147,0.006292379,0.0033868076,0.0016058321,0.005202002],"category_scores_gemma":[0.041939795,0.0014462142,0.004388218,0.0033945115,0.0046611875,0.012561506,0.007018921,0.006038729,0.0017282927],"study_design_candidate":"simulation_or_modeling","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":[0.00015539485,0.00014495544,0.004179864,0.00070769613,0.00012495606,0.00016118184,0.0010046522,0.06679002,0.0028821155,0.72031325,0.007593222,0.19594269],"study_design_scores_gemma":[0.000041549352,0.00009548967,0.00051283394,0.0002653432,0.000052946,0.00020806777,0.0002170134,0.50959265,0.007961671,0.45220315,0.028760327,0.00008896195],"about_ca_topic_score_codex":0.0028391988,"about_ca_topic_score_gemma":0.0025668251,"teacher_disagreement_score":0.013667611,"about_ca_system_score_codex":0.003124471,"about_ca_system_score_gemma":0.004133917,"threshold_uncertainty_score":0.072282135},"labels":[],"label_agreement":null},{"id":"W4320024115","doi":"10.1109/bigdata55660.2022.10020387","title":"An Intelligent Class: The Development Of A Novel Context Capturing Method For The Functional Auto Classification Of Records","year":2022,"lang":"en","type":"article","venue":"2022 IEEE International Conference on Big Data (Big Data)","topic":"Digital and Traditional Archives Management","field":"Arts and Humanities","cited_by":1,"is_retracted":false,"has_abstract":true,"route_ca_aff":true,"route_ca_fund":false,"route_ca_venue":false,"route_about_ca":false,"ca_institutions":"University of British Columbia","funders":"","keywords":"Computer science; Classifier (UML); Artificial intelligence; Context (archaeology); Machine learning; Process (computing); Statistical classification; Set (abstract data type); One-class classification; Task (project management); Data mining; Engineering","score_opus":0.5787873684648848,"score_gpt":0.3602816516620805,"score_spread":0.21850571680280428,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W4320024115","genre_codex":"methods","genre_gemma":"methods","domain_codex":null,"domain_gemma":null,"model_version":"metacan-v3-hybrid-931329e0061c","genre_candidate":"methods","genre_consensus":"methods","domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.038100645,0.00043521018,0.94941473,0.00066755427,0.00029060282,0.00034193596,0.0005620894,0.0049848626,0.005202341],"genre_scores_gemma":[0.22757877,0.00030902325,0.76143473,0.0005278261,0.00023723264,0.0003492797,0.00097177434,0.0003543165,0.008236966],"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","domain_scores_codex":[0.99854916,0.00017235288,0.000097333235,0.0005372779,0.0005309103,0.000112926406],"domain_scores_gemma":[0.9983652,0.00046342047,0.00019855029,0.0003297174,0.00054551393,0.00009760083],"candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0011395448,0.00058862194,0.00078025146,0.0019599118,0.0009808326,0.0018727437,0.0016248224,0.0010090643,0.0023690232],"category_scores_gemma":[0.0038959957,0.00032128918,0.00079202873,0.001585445,0.00053830823,0.0026370287,0.0014645876,0.0016677204,0.0013503111],"study_design_candidate":"simulation_or_modeling","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":[0.0002517816,0.0002477051,0.009230618,0.00014025629,0.00006019934,0.00016807752,0.0008191464,0.005475278,0.022057487,0.008737153,0.008674349,0.94413793],"study_design_scores_gemma":[0.000076209435,0.00036155595,0.016219882,0.00013192168,0.00013442141,0.000897837,0.0009912517,0.8634073,0.04604265,0.018369375,0.05323703,0.000130618],"about_ca_topic_score_codex":0.0046972553,"about_ca_topic_score_gemma":0.007860785,"teacher_disagreement_score":0.0046972553,"about_ca_system_score_codex":0.0007486302,"about_ca_system_score_gemma":0.001378255,"threshold_uncertainty_score":0.009339869},"labels":[],"label_agreement":null},{"id":"W4320024117","doi":"10.1109/bigdata55660.2022.10021046","title":"A novel self-adaptive method for improving patient monitoring with composite early-warning scores","year":2022,"lang":"en","type":"article","venue":"2022 IEEE International Conference on Big Data (Big Data)","topic":"Healthcare Technology and Patient Monitoring","field":"Medicine","cited_by":1,"is_retracted":false,"has_abstract":true,"route_ca_aff":true,"route_ca_fund":false,"route_ca_venue":false,"route_about_ca":false,"ca_institutions":"University of Waterloo","funders":"","keywords":"Computer science; Wearable computer; Remote patient monitoring; Real-time computing; Pruning; Wearable technology; Transmission (telecommunications); Adaptive sampling; Machine learning; Artificial intelligence; Data mining; Embedded system; Medicine; Telecommunications","score_opus":0.3307912242688666,"score_gpt":0.395012270814312,"score_spread":0.06422104654544541,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W4320024117","genre_codex":"methods","genre_gemma":"methods","domain_codex":null,"domain_gemma":null,"model_version":"metacan-v3-hybrid-931329e0061c","genre_candidate":"methods","genre_consensus":"methods","domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.045254167,0.00042148907,0.9505029,0.00020005727,0.00015475662,0.00015176929,0.0002374235,0.0020040309,0.0010735355],"genre_scores_gemma":[0.44574887,0.0004597574,0.5482261,0.00042707677,0.00027348346,0.00029448263,0.00072511914,0.00022517287,0.0036199505],"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","domain_scores_codex":[0.9991937,0.00011655093,0.00007951752,0.00026878173,0.00027755325,0.000063805346],"domain_scores_gemma":[0.99878055,0.0004388799,0.00014409881,0.00011473525,0.0004465853,0.00007511853],"candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.00084455183,0.000909698,0.00094964507,0.0014431814,0.00025528934,0.0007314538,0.000972738,0.0007078245,0.0012293514],"category_scores_gemma":[0.0038614776,0.00031546486,0.0008179776,0.0010777679,0.0002384337,0.0010974418,0.00079366035,0.0009402243,0.00065696146],"study_design_candidate":"simulation_or_modeling","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":[0.00056532794,0.0005445454,0.010637231,0.00015057268,0.00013143584,0.00021570559,0.00022019738,0.027415883,0.044725984,0.0013133103,0.0043427907,0.90973705],"study_design_scores_gemma":[0.00007120831,0.00042655246,0.0117729725,0.000030444746,0.00013065276,0.00051608676,0.00006283009,0.95839924,0.021930559,0.0018779805,0.004719302,0.00006217632],"about_ca_topic_score_codex":0.0016509183,"about_ca_topic_score_gemma":0.0019694234,"teacher_disagreement_score":0.0016509183,"about_ca_system_score_codex":0.000247795,"about_ca_system_score_gemma":0.0005265867,"threshold_uncertainty_score":0.004466474},"labels":[],"label_agreement":null},{"id":"W4320024124","doi":"10.1109/bigdata55660.2022.10020422","title":"Dynamic Graph Summarization: Optimal and Scalable","year":2022,"lang":"en","type":"article","venue":"2022 IEEE International Conference on Big Data (Big Data)","topic":"Graph Theory and Algorithms","field":"Computer Science","cited_by":3,"is_retracted":false,"has_abstract":true,"route_ca_aff":true,"route_ca_fund":false,"route_ca_venue":false,"route_about_ca":false,"ca_institutions":"University of Victoria","funders":"","keywords":"Automatic summarization; Computer science; Scalability; Lossless compression; Graph; Theoretical computer science; Algorithm; Artificial intelligence; Data compression","score_opus":0.15820750215485105,"score_gpt":0.3135448753019634,"score_spread":0.15533737314711235,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W4320024124","genre_codex":"methods","genre_gemma":"methods","domain_codex":null,"domain_gemma":null,"model_version":"metacan-v3-hybrid-931329e0061c","genre_candidate":"methods","genre_consensus":"methods","domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.023124909,0.0014626987,0.9536057,0.00065981736,0.00020506953,0.00028163657,0.0015949758,0.015818741,0.0032464156],"genre_scores_gemma":[0.17170508,0.0006688097,0.81123304,0.0003198661,0.00023080186,0.00023869304,0.008927389,0.0014635938,0.0052127526],"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","domain_scores_codex":[0.9985663,0.00019751802,0.00010719541,0.00050955894,0.0005012594,0.000118069445],"domain_scores_gemma":[0.9969386,0.0008733531,0.00032940236,0.0010732284,0.0006387477,0.00014666615],"candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.00088213175,0.0017030224,0.0012322002,0.0026143368,0.00088415085,0.001901187,0.0022094208,0.0012520631,0.003004466],"category_scores_gemma":[0.006183981,0.0006758538,0.00077966653,0.0028569293,0.0006424216,0.004319291,0.0017336154,0.0013520927,0.0019893649],"study_design_candidate":"simulation_or_modeling","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":[0.0004363509,0.00023937711,0.0014273502,0.0005648188,0.00013129924,0.000265557,0.00038723752,0.12499929,0.040473033,0.011786106,0.048517,0.7707726],"study_design_scores_gemma":[0.0001423235,0.000272836,0.0012197008,0.000045536854,0.00011443433,0.00055485877,0.00035851102,0.8907195,0.036271766,0.043233953,0.027003428,0.00006319459],"about_ca_topic_score_codex":0.003365779,"about_ca_topic_score_gemma":0.0068220836,"teacher_disagreement_score":0.003365779,"about_ca_system_score_codex":0.0009780048,"about_ca_system_score_gemma":0.0011650588,"threshold_uncertainty_score":0.010050952},"labels":[],"label_agreement":null},{"id":"W4320024146","doi":"10.1109/bigdata55660.2022.10020239","title":"A Stylized Presence Detection System in the Era of Blockchain and Big Data","year":2022,"lang":"en","type":"article","venue":"2022 IEEE International Conference on Big Data (Big Data)","topic":"Bluetooth and Wireless Communication Technologies","field":"Computer Science","cited_by":1,"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":"Stylized fact; Big data; Computer science; Blockchain; Popularity; Beacon; Identification (biology); Internet of Things; Bluetooth; Data science; Computer security; Telecommunications; Data mining; Wireless","score_opus":0.2687715235562875,"score_gpt":0.32970328184096914,"score_spread":0.06093175828468167,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W4320024146","genre_codex":"methods","genre_gemma":"empirical","domain_codex":null,"domain_gemma":null,"model_version":"metacan-v3-hybrid-931329e0061c","genre_candidate":"empirical","genre_consensus":null,"domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.08045241,0.0004983822,0.89010304,0.002838904,0.00038992125,0.00052021514,0.00059643696,0.0123251,0.012275526],"genre_scores_gemma":[0.8758192,0.00021695755,0.11388951,0.00050840166,0.0001555787,0.00029423318,0.00053820224,0.00018018999,0.008397681],"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","domain_scores_codex":[0.9972715,0.0009131682,0.00020805067,0.00053713087,0.00082326296,0.0002468322],"domain_scores_gemma":[0.99313074,0.0018404352,0.00069789385,0.0025014502,0.0011618697,0.00066756847],"candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.00334486,0.00041858913,0.0009659063,0.0012964766,0.0015394686,0.0030299155,0.0017029982,0.0016622712,0.00365429],"category_scores_gemma":[0.009589803,0.00043184677,0.00035135768,0.0011059231,0.0018162213,0.0069819386,0.00508178,0.0017296819,0.001842333],"study_design_candidate":"simulation_or_modeling","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":[0.0024599119,0.0006303397,0.012446537,0.00047122643,0.00015096446,0.0021580914,0.0028285086,0.066804454,0.04706989,0.41296956,0.0269348,0.42507565],"study_design_scores_gemma":[0.00011534083,0.0003270654,0.0010795996,0.00007133393,0.000049041126,0.0006801421,0.0002453186,0.8257682,0.024078943,0.11614321,0.031303983,0.00013791159],"about_ca_topic_score_codex":0.0019193746,"about_ca_topic_score_gemma":0.0016054291,"teacher_disagreement_score":0.00365429,"about_ca_system_score_codex":0.0010332617,"about_ca_system_score_gemma":0.0018911355,"threshold_uncertainty_score":0.017689526},"labels":[],"label_agreement":null},{"id":"W4320024165","doi":"10.1109/bigdata55660.2022.10020320","title":"Preserving Privacy Integration and Mining for Big Temporal Co-occurrence Patterns","year":2022,"lang":"en","type":"article","venue":"2022 IEEE International Conference on Big Data (Big Data)","topic":"Privacy-Preserving Technologies in Data","field":"Computer Science","cited_by":7,"is_retracted":false,"has_abstract":true,"route_ca_aff":true,"route_ca_fund":true,"route_ca_venue":false,"route_about_ca":false,"ca_institutions":"University of Manitoba","funders":"University of Manitoba","keywords":"Transparency (behavior); Big data; Information privacy; Computer science; Internet privacy; Data integration; Privacy law; Hierarchy; Open data; Computer security; Data science; Privacy policy; Data mining; World Wide Web; Political science","score_opus":0.38674126848575496,"score_gpt":0.38499527520446797,"score_spread":0.001745993281286995,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W4320024165","genre_codex":"methods","genre_gemma":"empirical","domain_codex":null,"domain_gemma":null,"model_version":"metacan-v3-hybrid-931329e0061c","genre_candidate":"empirical","genre_consensus":null,"domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.11781513,0.00093522586,0.8713895,0.001568032,0.00008023129,0.00028519976,0.0040678307,0.0018504226,0.0020083971],"genre_scores_gemma":[0.6058532,0.0005190584,0.38113418,0.0004100922,0.00013652485,0.00026330276,0.010166509,0.0001536654,0.0013635206],"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","domain_scores_codex":[0.987518,0.0026599918,0.0017331048,0.0032658465,0.0039663594,0.000856732],"domain_scores_gemma":[0.97281873,0.008427052,0.0037271362,0.012417852,0.0020376076,0.00057172065],"candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0065139737,0.00086646137,0.0013818712,0.0031734246,0.0020803637,0.0040008673,0.0023115405,0.0013652783,0.00062655803],"category_scores_gemma":[0.02919683,0.00062668615,0.0024009976,0.0072893775,0.0016554546,0.008381825,0.0043830336,0.0021281454,0.0004844019],"study_design_candidate":"simulation_or_modeling","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":[0.0016347512,0.0011127795,0.120532736,0.0009510055,0.00084122346,0.0020447013,0.0040803473,0.11479949,0.029348642,0.11353904,0.014398563,0.5967168],"study_design_scores_gemma":[0.000049341386,0.00015954534,0.010751599,0.00008301705,0.0001768782,0.0014369232,0.0014733849,0.7745705,0.021991571,0.17563272,0.013613179,0.00006133097],"about_ca_topic_score_codex":0.0050228797,"about_ca_topic_score_gemma":0.007413277,"teacher_disagreement_score":0.0065139737,"about_ca_system_score_codex":0.0014187125,"about_ca_system_score_gemma":0.003848201,"threshold_uncertainty_score":0.034449577},"labels":[],"label_agreement":null},{"id":"W4320024183","doi":"10.1109/bigdata55660.2022.10020511","title":"Dynamic Ensemble Size Adjustment for Memory Constrained Mondrian Forest","year":2022,"lang":"en","type":"article","venue":"2022 IEEE International Conference on Big Data (Big Data)","topic":"Data Stream Mining 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":"Concordia University","funders":"","keywords":"Mondrian; Overfitting; Computer science; Machine learning; Ensemble learning; Tree (set theory); Data stream mining; Artificial intelligence; Memory model; Data mining; Shared memory; Parallel computing; Mathematics; Artificial neural network","score_opus":0.21622871598344806,"score_gpt":0.35008913582646745,"score_spread":0.13386041984301938,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W4320024183","genre_codex":"methods","genre_gemma":"empirical","domain_codex":null,"domain_gemma":null,"model_version":"metacan-v3-hybrid-931329e0061c","genre_candidate":"empirical","genre_consensus":null,"domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.07279263,0.0007060293,0.92212915,0.0002889073,0.00016653816,0.0001015424,0.00018208023,0.0019331538,0.0016999971],"genre_scores_gemma":[0.5541571,0.00025219648,0.44218007,0.000345728,0.00016529893,0.00026644548,0.0007754272,0.00033430077,0.0015233549],"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","domain_scores_codex":[0.9989504,0.00026176154,0.00007605456,0.00028520857,0.00025832877,0.00016821145],"domain_scores_gemma":[0.9959155,0.0020098817,0.00023708599,0.00059950416,0.0010419117,0.00019609847],"candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0037534605,0.0007728861,0.0012485856,0.0012031421,0.0012802193,0.0010046284,0.002191028,0.0014419237,0.0015900203],"category_scores_gemma":[0.01286405,0.00038037793,0.00080874184,0.00074929825,0.0005959302,0.002175404,0.0013823898,0.0016015078,0.00055485766],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_system_candidate":false,"about_ca_system_consensus":false,"study_design_scores_codex":[0.0005114623,0.00018269753,0.008136653,0.00010945049,0.000120917604,0.00015441605,0.00022372276,0.54675883,0.007358931,0.010162318,0.0088355355,0.41744506],"study_design_scores_gemma":[0.000012166602,0.00001998904,0.00035226572,0.000010205596,0.0000098909795,0.000037813523,0.000014922521,0.9940084,0.0012521874,0.0036744862,0.0005997181,0.000007986261],"about_ca_topic_score_codex":0.0051548574,"about_ca_topic_score_gemma":0.009597021,"teacher_disagreement_score":0.0051548574,"about_ca_system_score_codex":0.0009103406,"about_ca_system_score_gemma":0.0016617206,"threshold_uncertainty_score":0.019850373},"labels":[],"label_agreement":null},{"id":"W4320024230","doi":"10.1109/bigdata55660.2022.10020333","title":"Thick Data Techniques for Identifying Abnormality in Video Frames for Wireless Capsule Endoscopy","year":2022,"lang":"en","type":"article","venue":"2022 IEEE International Conference on Big Data (Big Data)","topic":"Gastrointestinal Bleeding Diagnosis and Treatment","field":"Medicine","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":"NOSM University; Lakehead University","funders":"","keywords":"Computer science; Artificial intelligence; Capsule endoscopy; Abnormality; Modality (human–computer interaction); Filter (signal processing); Decision tree; Deep learning; Computer vision; Feature (linguistics); Feature extraction; Pattern recognition (psychology); Medicine; Radiology","score_opus":0.4424567085698399,"score_gpt":0.42876867080501857,"score_spread":0.013688037764821304,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W4320024230","genre_codex":"methods","genre_gemma":"empirical","domain_codex":null,"domain_gemma":null,"model_version":"metacan-v3-hybrid-931329e0061c","genre_candidate":"empirical","genre_consensus":null,"domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.03202165,0.0019259691,0.9601663,0.0005202111,0.00037863854,0.00019299462,0.0013307723,0.0015334026,0.0019301331],"genre_scores_gemma":[0.24199277,0.00225419,0.74590456,0.00037307266,0.00038094458,0.00026921727,0.004016993,0.00027238735,0.004535842],"study_design_codex":"design_other","study_design_gemma":"bench_or_experimental","domain_scores_codex":[0.9991678,0.00012337296,0.000095144234,0.00016309509,0.00038411323,0.000066485],"domain_scores_gemma":[0.9976776,0.0006010651,0.00034606148,0.0004584233,0.00082652684,0.000090248075],"candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.00093957636,0.0015250928,0.0008050377,0.0025176278,0.00033813334,0.0012594878,0.0010059869,0.0011356014,0.002621608],"category_scores_gemma":[0.0047241994,0.00034964937,0.0008122787,0.0016822466,0.0005141721,0.0020736824,0.0014365537,0.0015354282,0.002038891],"study_design_candidate":"bench_or_experimental","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":[0.00070017524,0.00019491935,0.005645827,0.0003753756,0.0001235264,0.0005474265,0.00014862379,0.03420948,0.10568158,0.003903456,0.0088659385,0.83960366],"study_design_scores_gemma":[0.00004393686,0.00062661216,0.011715256,0.00018685448,0.00012565783,0.0011712044,0.0003272525,0.84551567,0.10060455,0.009470291,0.030088803,0.00012387932],"about_ca_topic_score_codex":0.0024121031,"about_ca_topic_score_gemma":0.0035472114,"teacher_disagreement_score":0.002621608,"about_ca_system_score_codex":0.0005182775,"about_ca_system_score_gemma":0.00050833356,"threshold_uncertainty_score":0.008770168},"labels":[],"label_agreement":null},{"id":"W4320024352","doi":"10.1109/bigdata55660.2022.10020258","title":"Review of Publically Available Health Big Data Sets","year":2022,"lang":"en","type":"article","venue":"2022 IEEE International Conference on Big Data (Big Data)","topic":"Data Quality and Management","field":"Decision Sciences","cited_by":4,"is_retracted":false,"has_abstract":true,"route_ca_aff":true,"route_ca_fund":true,"route_ca_venue":false,"route_about_ca":true,"ca_institutions":"University of Victoria","funders":"University of Victoria","keywords":"Health informatics; Public health informatics; Computer science; Informatics; Usability; Data science; Dashboard; Data mining; Information retrieval; World Wide Web; Public health; Medicine; Health policy; International health; Engineering","score_opus":0.8077064389291911,"score_gpt":0.5061207953285485,"score_spread":0.30158564360064266,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W4320024352","genre_codex":"review","genre_gemma":"review","domain_codex":null,"domain_gemma":null,"model_version":"metacan-v3-hybrid-931329e0061c","genre_candidate":"review","genre_consensus":"review","domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.013967888,0.7586977,0.015162415,0.036163013,0.0034144418,0.00093259965,0.14679027,0.0012141824,0.023657572],"genre_scores_gemma":[0.069992855,0.674496,0.033676088,0.011588323,0.0028793397,0.0018150486,0.20275864,0.00088221836,0.0019116073],"study_design_codex":"design_other","study_design_gemma":"not_applicable","domain_scores_codex":[0.9697468,0.008623609,0.006593123,0.0025926472,0.01149096,0.0009527735],"domain_scores_gemma":[0.6788823,0.23960556,0.017606607,0.017650329,0.043102138,0.003153044],"candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.031556513,0.0010528865,0.0022519093,0.051593013,0.0017359986,0.005988922,0.004058817,0.0016072616,0.0057794875],"category_scores_gemma":[0.17177486,0.0014325803,0.002924943,0.055131003,0.002227793,0.007366271,0.005693455,0.0025081665,0.0015409369],"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":[0.00034140688,0.000120888064,0.023355603,0.10689879,0.0016826637,0.0005018684,0.002406664,0.002688971,0.0012067872,0.03216124,0.26721948,0.5614157],"study_design_scores_gemma":[0.000032624925,0.000066475746,0.027799856,0.062882505,0.0006769891,0.0005831387,0.0020372183,0.0009506827,0.0010901388,0.007728277,0.89600104,0.00015104616],"about_ca_topic_score_codex":0.015418398,"about_ca_topic_score_gemma":0.025956936,"teacher_disagreement_score":0.051593013,"about_ca_system_score_codex":0.0043702116,"about_ca_system_score_gemma":0.018358639,"threshold_uncertainty_score":0.16688883},"labels":[],"label_agreement":null}]}