{"meta":{"query_hash":"5d0a88660265","filters":{"venue":"Proceedings of the 2022 International Conference on Management of Data"},"cohort_total":5,"direct_labels_cover":0,"predictions_cover":5,"exported":5,"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/5d0a88660265","api":"https://metacan.xera.ac/api/v1/cohort?venue=Proceedings+of+the+2022+International+Conference+on+Management+of+Data"},"results":[{"id":"W4281747838","doi":"10.1145/3514221.3526161","title":"Natto: Providing Distributed Transaction Prioritization for High-Contention Workloads","year":2022,"lang":"en","type":"article","venue":"Proceedings of the 2022 International Conference on Management of Data","topic":"Distributed systems and fault tolerance","field":"Computer Science","cited_by":9,"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; Database transaction; Timestamp; Distributed transaction; Transaction processing; Latency (audio); Prioritization; Online transaction processing; Distributed database; Database; Distributed computing; Computer network; Business; Telecommunications","score_opus":0.062399477194828606,"score_gpt":0.29384535846480053,"score_spread":0.23144588126997193,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W4281747838","genre_codex":"methods","genre_gemma":"empirical","domain_codex":null,"domain_gemma":null,"model_version":"codex-gemma-dda1882f352a","genre_candidate":"empirical","genre_consensus":null,"domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.02421271,0.000078517,0.9536161,0.008746656,0.002747063,0.0020090244,0.0033478304,0.00011336655,0.0051287455],"genre_scores_gemma":[0.9953728,0.000025070247,0.00350166,0.00003480684,0.000028957453,0.00013737568,0.00041139487,0.000006511799,0.00048142322],"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"simulation_or_modeling","domain_scores_codex":[0.9983574,0.000012980611,0.00037355212,0.0003990243,0.00073044543,0.00012655572],"domain_scores_gemma":[0.9988031,0.000021529753,0.00047989268,0.00032196354,0.00035424464,0.000019269186],"candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0005920526,0.00010539723,0.00013985655,0.00009526395,0.00017709004,0.00011220959,0.0031896126,0.00002100517,0.00003335876],"category_scores_gemma":[0.000028522227,0.00009544691,0.000059243375,0.00037884386,0.000026498792,0.0008354929,0.00091993576,0.00011230375,5.5758926e-7],"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":[0.000102673264,0.0002056395,0.0002658563,0.00020665325,0.00015273411,3.4841597e-7,0.00008062709,0.0003239882,0.0037905027,0.97326285,0.0037279825,0.017880166],"study_design_scores_gemma":[0.0049717207,0.000781646,0.010622644,0.0016226812,0.00023159412,0.00001262441,0.0054287515,0.8504345,0.012912813,0.07250894,0.039624047,0.00084802683],"about_ca_topic_score_codex":0.00001104189,"about_ca_topic_score_gemma":6.44126e-7,"teacher_disagreement_score":0.9711601,"about_ca_system_score_codex":0.00009324665,"about_ca_system_score_gemma":0.00002298212,"threshold_uncertainty_score":0.59271467},"labels":[],"label_agreement":null},{"id":"W4281775561","doi":"10.1145/3514221.3517834","title":"Proteus: Autonomous Adaptive Storage for Mixed Workloads","year":2022,"lang":"en","type":"article","venue":"Proceedings of the 2022 International Conference on Management of Data","topic":"Cloud Computing and Resource Management","field":"Computer Science","cited_by":17,"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":"Online transaction processing; Online analytical processing; Computer science; Workload; Distributed transaction; Transaction processing; Database; Database transaction; Transaction processing system; Operating system; Distributed computing; Data warehouse","score_opus":0.07251230446442274,"score_gpt":0.2891402640180174,"score_spread":0.21662795955359465,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W4281775561","genre_codex":"other","genre_gemma":"empirical","domain_codex":null,"domain_gemma":null,"model_version":"codex-gemma-dda1882f352a","genre_candidate":"empirical","genre_consensus":null,"domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.23478238,0.00044312276,0.23635948,0.09374557,0.011597511,0.014513926,0.0027489308,0.0009039417,0.40490514],"genre_scores_gemma":[0.9751356,0.000009909634,0.019981487,0.00014108354,0.000049413127,0.00019285615,0.000021767428,0.000011373844,0.004456494],"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"simulation_or_modeling","domain_scores_codex":[0.99809134,0.000017085094,0.00032654384,0.0005267138,0.0008626721,0.00017563636],"domain_scores_gemma":[0.9986655,0.000040528495,0.00048107535,0.00059465185,0.0001917907,0.000026468519],"candidate_categories":["open_science"],"consensus_categories":["open_science"],"category_scores_codex":[0.00086913223,0.00013621792,0.00015990646,0.00015344535,0.00020007546,0.00008774382,0.007959985,0.00001565572,0.000054784592],"category_scores_gemma":[0.000027864926,0.00011461992,0.00008482329,0.00032820445,0.00005303166,0.000105598876,0.01011094,0.00016538966,0.0000018927071],"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":[0.00008168839,0.00029403,0.00005863875,0.0001085285,0.00031848924,0.0000011769065,0.00024346245,0.0018033416,0.0002888821,0.9357973,0.01525304,0.04575146],"study_design_scores_gemma":[0.0010141701,0.0005022001,0.0010760835,0.00025913856,0.00006902024,0.0000032398293,0.003092213,0.91478395,0.00064722373,0.02391801,0.054330695,0.00030406335],"about_ca_topic_score_codex":0.00000870833,"about_ca_topic_score_gemma":4.839826e-7,"teacher_disagreement_score":0.9129806,"about_ca_system_score_codex":0.00009203583,"about_ca_system_score_gemma":0.000028660508,"threshold_uncertainty_score":0.9978951},"labels":[],"label_agreement":null},{"id":"W4281779720","doi":"10.1145/3514221.3517837","title":"A Convex-Programming Approach for Efficient Directed Densest Subgraph Discovery","year":2022,"lang":"en","type":"article","venue":"Proceedings of the 2022 International Conference on Management of Data","topic":"Advanced Graph Neural Networks","field":"Computer Science","cited_by":29,"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":"University of Hong Kong","keywords":"Scalability; Parameterized complexity; Computer science; Bipartite graph; Approximation algorithm; Linear programming; Matching (statistics); Theoretical computer science; Induced subgraph isomorphism problem; Directed graph; Graph; Mathematical optimization; Algorithm; Mathematics; Line graph","score_opus":0.06422309090837273,"score_gpt":0.2962473222413728,"score_spread":0.23202423133300004,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W4281779720","genre_codex":"methods","genre_gemma":"empirical","domain_codex":null,"domain_gemma":null,"model_version":"codex-gemma-dda1882f352a","genre_candidate":"empirical","genre_consensus":null,"domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.10886915,0.00030405406,0.83398396,0.008358441,0.003830393,0.0065178657,0.0019757326,0.0005720554,0.03558838],"genre_scores_gemma":[0.9272701,0.000037686867,0.0715167,0.00010410518,0.000024277246,0.0002141792,0.00012433551,0.0000128432275,0.0006957473],"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"simulation_or_modeling","domain_scores_codex":[0.9979525,0.000013769331,0.0003336172,0.0006005149,0.000890944,0.00020862673],"domain_scores_gemma":[0.9986861,0.000053347805,0.00046525817,0.00054569653,0.00022087083,0.00002876585],"candidate_categories":["open_science"],"consensus_categories":[],"category_scores_codex":[0.00048040875,0.00014920908,0.00017964031,0.0001709663,0.00017867054,0.00012879717,0.006392463,0.000016935164,0.000014064316],"category_scores_gemma":[0.00003509528,0.00012486524,0.000098660916,0.0006215093,0.00009523641,0.0005453807,0.00543437,0.00018120672,2.7524354e-7],"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":[0.00012179789,0.00036793313,0.00051956926,0.00012902467,0.00019771847,7.890182e-7,0.000079993675,0.0017528184,0.0008750056,0.98288846,0.001835809,0.011231068],"study_design_scores_gemma":[0.0010132089,0.00026318536,0.0012315287,0.00013213044,0.000073310526,0.000007701024,0.0011172794,0.9761641,0.0013473628,0.0129002705,0.005416006,0.0003338885],"about_ca_topic_score_codex":0.000004721988,"about_ca_topic_score_gemma":3.4057257e-7,"teacher_disagreement_score":0.9744113,"about_ca_system_score_codex":0.000045152643,"about_ca_system_score_gemma":0.000017740229,"threshold_uncertainty_score":0.99898344},"labels":[],"label_agreement":null},{"id":"W4281788841","doi":"10.1145/3514221.3517904","title":"FILA: Online Auditing of Machine Learning Model Accuracy under Finite Labelling Budget","year":2022,"lang":"en","type":"article","venue":"Proceedings of the 2022 International Conference on Management of Data","topic":"Machine Learning and Data Classification","field":"Computer Science","cited_by":1,"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 Toronto","funders":"Natural Sciences and Engineering Research Council of Canada","keywords":"Audit; Pipeline (software); Computer science; Process (computing); Training set; Artificial intelligence; Machine learning; Data modeling; Labelling; Software engineering; Accounting; Programming language","score_opus":0.10600199055667883,"score_gpt":0.32951755703808333,"score_spread":0.2235155664814045,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W4281788841","genre_codex":"methods","genre_gemma":"empirical","domain_codex":null,"domain_gemma":null,"model_version":"codex-gemma-dda1882f352a","genre_candidate":"empirical","genre_consensus":null,"domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.18848358,0.00040515166,0.6243018,0.062584445,0.001852498,0.0019081505,0.0074581127,0.00045704492,0.11254924],"genre_scores_gemma":[0.97340477,0.00018176044,0.02429051,0.000114744704,0.000018207633,0.000011277699,0.00056434795,0.0000092881255,0.0014050844],"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"simulation_or_modeling","domain_scores_codex":[0.9981307,0.000028302478,0.0004255594,0.0004101492,0.00088053494,0.00012476086],"domain_scores_gemma":[0.9982321,0.00011112721,0.00092281855,0.0004820549,0.00023004205,0.00002184209],"candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0008681192,0.00011791876,0.0001576705,0.00017296795,0.00015695923,0.000058122496,0.0049663987,0.00001695986,0.00013322014],"category_scores_gemma":[0.00022749406,0.000104081286,0.00004749417,0.00034704932,0.00004433669,0.00059223466,0.0054974426,0.00037283002,0.0000015787123],"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":[0.00005609847,0.00034248238,0.001718706,0.00020025292,0.00017721608,5.0093325e-7,0.00023733161,0.06650443,0.0047902907,0.90249616,0.0011494854,0.022327073],"study_design_scores_gemma":[0.00026115403,0.00005920517,0.0007473597,0.00010923607,0.000022682303,0.0000010599582,0.000502049,0.989136,0.00037588325,0.006795807,0.0018956711,0.00009391975],"about_ca_topic_score_codex":0.000031593365,"about_ca_topic_score_gemma":0.000001619123,"teacher_disagreement_score":0.92263156,"about_ca_system_score_codex":0.000037589427,"about_ca_system_score_gemma":0.000034603505,"threshold_uncertainty_score":0.92288876},"labels":[],"label_agreement":null},{"id":"W4282597468","doi":"10.1145/3514221.3517900","title":"One Size Does Not Fit All: A Bandit-Based Sampler Combination Framework with Theoretical Guarantees","year":2022,"lang":"en","type":"article","venue":"Proceedings of the 2022 International Conference on Management of Data","topic":"Data Stream Mining Techniques","field":"Computer Science","cited_by":7,"is_retracted":false,"has_abstract":true,"route_ca_aff":true,"route_ca_fund":false,"route_ca_venue":false,"route_about_ca":false,"ca_institutions":"Simon Fraser University","funders":"","keywords":"Computer science; Sample (material); Population; Set (abstract data type); Sampling (signal processing); Measure (data warehouse); Data mining; Mathematical optimization; Mathematics","score_opus":0.07058486992079954,"score_gpt":0.31386200692866906,"score_spread":0.24327713700786952,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W4282597468","genre_codex":"methods","genre_gemma":"empirical","domain_codex":null,"domain_gemma":null,"model_version":"codex-gemma-dda1882f352a","genre_candidate":"empirical","genre_consensus":null,"domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.2372356,0.000044756016,0.42165548,0.21973184,0.0026482125,0.0056406204,0.010764245,0.001386132,0.10089311],"genre_scores_gemma":[0.8915988,0.000018296934,0.107573986,0.00048184517,0.000012584368,0.000074782736,0.000092099974,0.000010104199,0.00013750397],"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"theoretical_or_conceptual","domain_scores_codex":[0.99755317,0.00002485331,0.00030744544,0.0005197566,0.0014376031,0.00015716076],"domain_scores_gemma":[0.99835694,0.00018626907,0.00040702237,0.0007896041,0.00023410414,0.000026079073],"candidate_categories":["open_science"],"consensus_categories":[],"category_scores_codex":[0.000768778,0.00014846213,0.00017907115,0.00013402344,0.00010225999,0.00014960967,0.007705489,0.000025955942,0.00055921386],"category_scores_gemma":[0.00015830282,0.000104659775,0.000042395066,0.00029400716,0.00019621268,0.00055656774,0.0048900396,0.00025880785,0.0000014453332],"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":[0.0001569912,0.00037304242,0.00039444963,0.000058726713,0.00015473277,0.0000010143577,0.0000862914,0.0000072193047,0.00052992173,0.9930882,0.001874222,0.0032751926],"study_design_scores_gemma":[0.0023751077,0.0017423184,0.009483857,0.0015985171,0.00024555222,0.000007330477,0.0012152501,0.15472487,0.032084275,0.7883603,0.0073593743,0.0008032326],"about_ca_topic_score_codex":0.000013525914,"about_ca_topic_score_gemma":9.564192e-7,"teacher_disagreement_score":0.6543632,"about_ca_system_score_codex":0.000059627873,"about_ca_system_score_gemma":0.000033743832,"threshold_uncertainty_score":0.9976633},"labels":[],"label_agreement":null}]}