{"meta":{"page":1,"per_page":50,"max_per_page":100,"total":4,"total_is_capped":false,"direct_labels_cover":0,"predictions_cover":4,"direct_label_status":"direct model label, unvalidated","prediction_status":"machine_predicted_unvalidated (Codex and Gemma teacher distillation)","score_status":"score_only:v0-immature-baseline (scores rank; they never assert a category)","snapshot":{"source":"OpenAlex, pinned release, all 482 partitions","release":"2026-06-24","frame_built":"2026-07-12","author_layer_release":"2026-06-26"},"query_hash":"382c3c355343","filters":{"venue":"2016 SAI Computing Conference (SAI)"}},"results":[{"id":"W2513414733","doi":"10.1109/sai.2016.7556026","title":"A density-based data cleaning approach for deduplication with data consistency and accuracy","year":2016,"lang":"en","type":"article","venue":"2016 SAI Computing Conference (SAI)","topic":"Data Quality and Management","field":"Decision Sciences","cited_by":11,"is_retracted":false,"has_abstract":true,"routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false},"ca_institutions":"McMaster University","funders":"","keywords":"Data deduplication; Computer science; Data mining; Correctness; Data warehouse; Data quality; Tuple; Data transformation; Consistency (knowledge bases); Scalability; Data consistency; Data modeling; Data set; Database; Data integrity; Set (abstract data type); Algorithm; Artificial intelligence; Mathematics","authors":[{"name":"Samir Al-Janabi","is_ca":true},{"name":"Ryszard Janicki","is_ca":true}],"retraction":null,"screen_n_in":null,"score":{"opus":0.4115022651312117,"gpt":0.4283854050917017,"spread":0.01688313996049001,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.005960433,0.0002351657,0.0003900812,0.0001615827,0.0004224492,0.0007973299,0.003572036,0.00007366818,0.00003745212],"category_scores_gemma":[0.005022718,0.0001411539,0.00002763475,0.0002874612,0.0003282972,0.001227319,0.002670178,0.000105327,0.00007314172],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00002700786,"about_ca_system_score_gemma":0.0003694673,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0001205973,"about_ca_topic_score_gemma":0.00008011714,"domain_scores_codex":[0.996304,0.0002699591,0.0006547307,0.001625794,0.0007412083,0.0004042967],"domain_scores_gemma":[0.9903907,0.00386167,0.0005871859,0.004487969,0.0005115604,0.0001608689],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0001546884,0.0001567576,0.004656238,0.00008053494,0.00006318272,0.000002874186,0.0002013646,0.00002511128,0.0003346727,0.04588323,0.0456972,0.9027441],"study_design_scores_gemma":[0.003913671,0.0002856824,0.014247,0.0006521974,0.0001903802,0.00002505865,0.001704682,0.7024488,0.0002643889,0.01036221,0.2648785,0.001027402],"study_design_candidate":"design_other","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.01451239,0.00005192398,0.9797456,0.002861469,0.00008974913,0.0006836994,0.0004597144,0.000102925,0.001492521],"genre_scores_gemma":[0.8668959,0.00001814706,0.1311161,0.0008169618,0.0001059993,0.000009748591,0.0007391997,0.00001693888,0.0002809842],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.9017168,"threshold_uncertainty_score":0.7688671,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null},{"id":"W2518601541","doi":"10.1109/sai.2016.7556142","title":"An analysis of live migration in openstack using high speed optical network","year":2016,"lang":"en","type":"article","venue":"2016 SAI Computing Conference (SAI)","topic":"Cloud Computing and Resource Management","field":"Computer Science","cited_by":7,"is_retracted":false,"has_abstract":true,"routes":{"ca_aff":false,"ca_fund":false,"ca_venue":false,"about_ca":true},"ca_institutions":"","funders":"","keywords":"Downtime; Live migration; Computer science; Virtual machine; Virtualization; Latency (audio); Operating system; Host (biology); Computer network; Cloud computing; Telecommunications","authors":[{"name":"Md Israfil Biswas","is_ca":false},{"name":"Gerard Parr","is_ca":false},{"name":"Sally McClean","is_ca":false},{"name":"Philip Morrow","is_ca":false},{"name":"Bryan Scotney","is_ca":false}],"retraction":null,"screen_n_in":null,"score":{"opus":0.02813250093470578,"gpt":0.2739620232304108,"spread":0.245829522295705,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.001116927,0.0002845648,0.0006265332,0.0004690928,0.0001577188,0.0002504274,0.001364428,0.0001215367,0.00003737098],"category_scores_gemma":[0.0000727318,0.0002133323,0.000157931,0.00139242,0.0001311153,0.0001219667,0.000724219,0.0001621517,0.00003394065],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001450516,"about_ca_system_score_gemma":0.0001412669,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001174305,"about_ca_topic_score_gemma":0.0002141378,"domain_scores_codex":[0.9970281,0.0003259387,0.0007265233,0.0008220877,0.0004649395,0.0006323991],"domain_scores_gemma":[0.9977847,0.0003925487,0.0003958766,0.0009996761,0.0002648802,0.0001623064],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.00004104313,0.0003798838,0.07522919,0.00003448176,0.0004525118,0.00004758005,0.002098125,0.6822027,0.003905721,0.07615516,0.0002445264,0.1592091],"study_design_scores_gemma":[0.0004646081,0.0001300431,0.0799254,0.0003568998,0.0001036096,0.000002642492,0.00007838506,0.9176913,0.0002306759,0.0005776337,0.0001292983,0.0003095061],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.6114219,0.00002073379,0.3874886,0.0003533484,0.0001817291,0.0001401555,0.000001272573,0.00009213549,0.0003000539],"genre_scores_gemma":[0.9656073,0.000007261325,0.03396196,0.0001049824,0.0001965869,8.602805e-7,0.000004304565,0.00001367918,0.0001031057],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.3541853,"threshold_uncertainty_score":0.869944,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null},{"id":"W2517886881","doi":"10.1109/sai.2016.7555963","title":"VENCE: A new machine learning method enhanced by ontological knowledge to extract summaries","year":2016,"lang":"en","type":"article","venue":"2016 SAI Computing Conference (SAI)","topic":"Text and Document Classification Technologies","field":"Computer Science","cited_by":2,"is_retracted":false,"has_abstract":true,"routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false},"ca_institutions":"Université Laval","funders":"","keywords":"Computer science; Ontology; Set (abstract data type); Jaccard index; Process (computing); Information retrieval; Function (biology); Domain (mathematical analysis); Artificial intelligence; Quality (philosophy); Space (punctuation); Natural language processing; Machine learning; Data mining; Pattern recognition (psychology); Mathematics","authors":[{"name":"Jésus Antonio Motta","is_ca":true},{"name":"Laurence Capus","is_ca":true},{"name":"Nicole Tourigny","is_ca":false}],"retraction":null,"screen_n_in":null,"score":{"opus":0.0349061431540296,"gpt":0.3186361008192224,"spread":0.2837299576651928,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":["metaepi_narrow","insufficient_payload"],"consensus_categories":[],"category_scores_codex":[0.0007337276,0.0003713737,0.000461173,0.0002092714,0.0003367544,0.0004738106,0.002095122,0.0002241757,0.000282193],"category_scores_gemma":[0.0009742161,0.0002553712,0.0001056204,0.0005275266,0.0001481214,0.0005325417,0.0009954788,0.0003781575,0.001194771],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001318188,"about_ca_system_score_gemma":0.0003136472,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0001095859,"about_ca_topic_score_gemma":0.00002725503,"domain_scores_codex":[0.9971327,0.0002820966,0.0005194873,0.001007219,0.0003422564,0.0007162649],"domain_scores_gemma":[0.9973902,0.0009886299,0.0002986518,0.0008173619,0.0002284917,0.0002767027],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"not_applicable","study_design_scores_codex":[0.00001524354,0.00007037018,0.0005350466,0.000008049287,0.00001723154,0.000002459419,0.0006261989,0.000008359497,0.03407032,0.06554335,0.01861284,0.8804905],"study_design_scores_gemma":[0.001701973,0.0008245721,0.003542555,0.0006459269,0.00002417106,0.00002060677,0.0001936346,0.03374564,0.1059588,0.0178147,0.834002,0.001525501],"study_design_candidate":"design_other","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.005063965,0.0004649174,0.9796009,0.006986099,0.0003795939,0.0002692191,0.00000560505,0.001535866,0.005693817],"genre_scores_gemma":[0.8241614,0.00009157649,0.1537343,0.0002041549,0.0001061683,0.00001786753,0.000006766891,0.00001855914,0.02165925],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.878965,"threshold_uncertainty_score":0.9999899,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null},{"id":"W2514676647","doi":"10.1109/sai.2016.7555968","title":"An overview of IRFS on Canadian GAAP — Self-organizing maps (SOMs)<sup>MS</sup>","year":2016,"lang":"en","type":"article","venue":"2016 SAI Computing Conference (SAI)","topic":"Financial Reporting and Valuation Research","field":"Business, Management and Accounting","cited_by":1,"is_retracted":false,"has_abstract":true,"routes":{"ca_aff":false,"ca_fund":false,"ca_venue":false,"about_ca":true},"ca_institutions":"","funders":"","keywords":"Accounting; Financial statement; International Financial Reporting Standards; Comparability; Business; Profitability index; Leverage (statistics); Accounting management; Financial accounting; Market liquidity; Accounting standard; Financial ratio; Accounting information system; Finance; Computer science; Audit","authors":[{"name":"M. Shanmuganathan","is_ca":false}],"retraction":null,"screen_n_in":null,"score":{"opus":0.09217291123567606,"gpt":0.3205029401260429,"spread":0.2283300288903669,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":["insufficient_payload"],"consensus_categories":[],"category_scores_codex":[0.001883469,0.0003077274,0.0004297261,0.0005276913,0.0004143598,0.0003535384,0.0006451366,0.0001613877,0.0007118583],"category_scores_gemma":[0.0009947184,0.0002394007,0.0001077236,0.0005905298,0.0000907975,0.0005494463,0.000190462,0.0002420026,0.001039698],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001739208,"about_ca_system_score_gemma":0.0008191545,"about_ca_topic_candidate":true,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.04227883,"about_ca_topic_score_gemma":0.003685648,"domain_scores_codex":[0.9971345,0.00008501112,0.0007004703,0.0006039909,0.0006840011,0.0007919926],"domain_scores_gemma":[0.9974946,0.0001884747,0.0005453821,0.0006505019,0.001003773,0.0001172913],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"not_applicable","study_design_scores_codex":[0.00008147179,0.0005284026,0.2389822,0.0009999154,0.0001583977,0.0001016176,0.0009374937,0.000505218,0.004665828,0.2548549,0.04393089,0.4542537],"study_design_scores_gemma":[0.003503606,0.0003796218,0.3161107,0.006265421,0.0001858207,0.00001960295,0.0007067787,0.0989332,0.001618477,0.01245295,0.5574498,0.002373978],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9569952,0.0001295203,0.001112329,0.002579561,0.0003856228,0.0005060855,0.00001461414,0.0003559494,0.03792107],"genre_scores_gemma":[0.9969724,0.00004734868,0.0004288045,0.0007647132,0.000946787,0.00000547744,0.00002683489,0.00005040194,0.0007572335],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.5135189,"threshold_uncertainty_score":0.9997381,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null}]}