{"meta":{"query_hash":"293684bd6cd1","filters":{"venue":"Foundations and Trends in Databases"},"cohort_total":4,"direct_labels_cover":0,"predictions_cover":4,"exported":4,"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/293684bd6cd1","api":"https://metacan.xera.ac/api/v1/cohort?venue=Foundations+and+Trends+in+Databases"},"results":[{"id":"W2791946429","doi":"10.1561/1900000055","title":"Algorithmic Aspects of Parallel Data Processing","year":2018,"lang":"en","type":"article","venue":"Foundations and Trends in Databases","topic":"Algorithms and Data Compression","field":"Computer Science","cited_by":24,"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; Data processing; Massively parallel; SPARK (programming language); Parallel processing; Sorting; Joins; Data processing system; Scope (computer science); Computation; Parallel computing; Distributed computing; Theoretical computer science; Algorithm; Database","score_opus":0.08973294613584182,"score_gpt":0.36032610517810243,"score_spread":0.2705931590422606,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W2791946429","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":[0.0051820506,0.009456567,0.953331,0.0034359077,0.0004245716,0.00016002204,0.00021263785,0.0003256033,0.027471775],"genre_scores_gemma":[0.18722944,0.024242045,0.7656401,0.0013180721,0.0032449532,0.0009916673,0.0013878592,0.00044269665,0.015503094],"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"theoretical_or_conceptual","domain_scores_codex":[0.99552995,0.0009777508,0.0003687476,0.00068041385,0.0022071977,0.00023606577],"domain_scores_gemma":[0.9958896,0.002379265,0.0001626448,0.00085244246,0.0006375653,0.00007859058],"candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0029281739,0.0013078781,0.0010732599,0.0016988858,0.0013050522,0.004685568,0.002427862,0.001362633,0.0043108137],"category_scores_gemma":[0.012358523,0.00085516216,0.0013338526,0.0042700283,0.0032820515,0.0067312866,0.0028948851,0.003504942,0.0018056682],"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":[0.000036731093,0.00004401791,0.0004923624,0.0004819036,0.000045616904,0.00009280739,0.00018978806,0.045272294,0.00064856245,0.87828416,0.0057624136,0.06864947],"study_design_scores_gemma":[0.00001452646,0.00001495915,0.00012293046,0.0000474176,0.000013541365,0.00010753869,0.000039662664,0.09399339,0.00047251864,0.88428736,0.02087571,0.000010532237],"about_ca_topic_score_codex":0.001283319,"about_ca_topic_score_gemma":0.0011589818,"teacher_disagreement_score":0.004685568,"about_ca_system_score_codex":0.0017217671,"about_ca_system_score_gemma":0.0020179823,"threshold_uncertainty_score":0.015485823},"labels":[],"label_agreement":null},{"id":"W4285522518","doi":"10.1561/1900000066","title":"Differential Privacy for Databases","year":2021,"lang":"en","type":"article","venue":"Foundations and Trends in Databases","topic":"Privacy-Preserving Technologies in Data","field":"Computer Science","cited_by":13,"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":"Differential privacy; Database; Computer science; Differential (mechanical device); Internet privacy; Data mining; Engineering","score_opus":0.09559784109037046,"score_gpt":0.35532375681894873,"score_spread":0.2597259157285783,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W4285522518","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":[0.00483003,0.04246464,0.83320135,0.016420443,0.0019071056,0.00019351568,0.0012268897,0.0007448161,0.09901122],"genre_scores_gemma":[0.2958387,0.08298925,0.51231676,0.010951456,0.0070739663,0.0010874303,0.0027940853,0.0006003591,0.08634796],"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"theoretical_or_conceptual","domain_scores_codex":[0.9953152,0.0012595966,0.0003539956,0.0007442851,0.0020523227,0.00027460884],"domain_scores_gemma":[0.99681073,0.001642283,0.00016968838,0.00079597556,0.00045489857,0.00012638162],"candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0031133755,0.0010003428,0.0010042484,0.001974787,0.0018716146,0.007342173,0.00154963,0.0022081146,0.0071095033],"category_scores_gemma":[0.009000429,0.0006095844,0.0014637846,0.004854387,0.0037264922,0.011053714,0.004133232,0.0061565824,0.002735642],"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":[0.0000097618,0.000007845249,0.00007797604,0.00010677055,0.000009974338,0.00004053009,0.00010153237,0.00098501,0.00025569234,0.9638285,0.009889005,0.024687426],"study_design_scores_gemma":[0.0000063813036,0.000013793623,0.000072839364,0.0000658367,0.000007983167,0.00016462764,0.00004107958,0.004726563,0.00035062034,0.8946359,0.099904165,0.000010166358],"about_ca_topic_score_codex":0.0014363478,"about_ca_topic_score_gemma":0.0007857505,"teacher_disagreement_score":0.007342173,"about_ca_system_score_codex":0.0029562241,"about_ca_system_score_gemma":0.001877679,"threshold_uncertainty_score":0.023783624},"labels":[],"label_agreement":null},{"id":"W4293582904","doi":"10.1561/1900000045","title":"Trends in Cleaning Relational Data: Consistency and Deduplication","year":2015,"lang":"en","type":"article","venue":"Foundations and Trends in Databases","topic":"Privacy-Preserving Technologies in Data","field":"Computer Science","cited_by":124,"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":"Data deduplication; Consistency (knowledge bases); Data consistency; Computer science; Database; Data science; Artificial intelligence","score_opus":0.20184294516867166,"score_gpt":0.36898336875878557,"score_spread":0.1671404235901139,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W4293582904","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":[0.030551871,0.31456158,0.5577621,0.06270677,0.0033799661,0.00044197717,0.0018704132,0.0023197513,0.026405534],"genre_scores_gemma":[0.23430309,0.25403798,0.46964082,0.016107393,0.008358362,0.00053102366,0.006538611,0.001286358,0.009196432],"study_design_codex":"design_other","study_design_gemma":"not_applicable","domain_scores_codex":[0.9747591,0.005694814,0.0019046371,0.0035451169,0.013349856,0.0007464622],"domain_scores_gemma":[0.92628324,0.034971014,0.0055028456,0.017008523,0.015027863,0.0012066052],"candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.019363683,0.0008981948,0.0014482768,0.005585612,0.0019483346,0.008655013,0.006440555,0.0029835608,0.0030666047],"category_scores_gemma":[0.05149089,0.0012736062,0.0014869573,0.014061909,0.00411055,0.024286594,0.0061778296,0.005761421,0.0017147289],"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":[0.0002429244,0.00018035264,0.013112881,0.0034122064,0.00015742028,0.00029040105,0.0016384494,0.008498482,0.0039790543,0.24906503,0.053155705,0.66626716],"study_design_scores_gemma":[0.00006272256,0.00025826402,0.00901033,0.0021933587,0.00016041945,0.0033793473,0.0027666534,0.05077135,0.01655716,0.262406,0.6522445,0.0001899801],"about_ca_topic_score_codex":0.0026300875,"about_ca_topic_score_gemma":0.0018443071,"teacher_disagreement_score":0.019363683,"about_ca_system_score_codex":0.0031440253,"about_ca_system_score_gemma":0.0032785626,"threshold_uncertainty_score":0.102406204},"labels":[],"label_agreement":null},{"id":"W4403690984","doi":"10.1561/1900000090","title":"Modern Techniques For Querying Graph-structured Databases","year":2024,"lang":"en","type":"article","venue":"Foundations and Trends in Databases","topic":"Graph Theory and Algorithms","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":"University of Waterloo; Polytechnique Montréal","funders":"","keywords":"Computer science; Graph database; Database; Graph; Information retrieval; Theoretical computer science","score_opus":0.048239545733152527,"score_gpt":0.3378664988143314,"score_spread":0.28962695308117886,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W4403690984","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":[0.0021521675,0.0039591608,0.98693395,0.0011538853,0.00010941919,0.0002231913,0.0011525294,0.0023921889,0.0019235184],"genre_scores_gemma":[0.043677427,0.009787101,0.9396218,0.0006515037,0.00032224928,0.00039182726,0.003769041,0.0004625456,0.0013164137],"study_design_codex":"design_other","study_design_gemma":"theoretical_or_conceptual","domain_scores_codex":[0.9927927,0.0015215444,0.00089046964,0.0010329377,0.003441348,0.00032091347],"domain_scores_gemma":[0.9917585,0.003621379,0.00047906075,0.0025854816,0.0014271047,0.00012843453],"candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0042251367,0.0015538919,0.0017747056,0.00536889,0.0012798547,0.005981127,0.005632863,0.0018509587,0.004362927],"category_scores_gemma":[0.017763693,0.0012476381,0.0023652767,0.012704536,0.0016885985,0.011576479,0.004429375,0.0030641379,0.0025199915],"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":[0.00023961466,0.00018149216,0.0017217156,0.002958516,0.00031487728,0.00024298854,0.0012069817,0.05793254,0.009144394,0.34209153,0.049683604,0.53428185],"study_design_scores_gemma":[0.00012020886,0.00012867167,0.00053589174,0.00034889768,0.0001491792,0.0008957183,0.0007465205,0.2899482,0.009001044,0.5924486,0.10558039,0.000096640746],"about_ca_topic_score_codex":0.0031782812,"about_ca_topic_score_gemma":0.004391304,"teacher_disagreement_score":0.005981127,"about_ca_system_score_codex":0.0017168827,"about_ca_system_score_gemma":0.0027111433,"threshold_uncertainty_score":0.022344887},"labels":[],"label_agreement":null}]}