{"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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.006211936,0.001255333,0.001870586,0.003812885,0.001875623,0.002869184,0.004685185,0.001537053,0.001349821],"category_scores_gemma":[0.03130373,0.001044863,0.001768924,0.004669575,0.00138042,0.005635941,0.004658213,0.002236048,0.0005436629],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.002032753,"about_ca_system_score_gemma":0.002612976,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.004585028,"about_ca_topic_score_gemma":0.00464976,"domain_scores_codex":[0.9928739,0.00142469,0.0006471683,0.001110992,0.0035502,0.0003930402],"domain_scores_gemma":[0.9826925,0.004717656,0.001733619,0.004319694,0.006078258,0.0004582465],"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":[0.0003447802,0.0004910741,0.01077938,0.000443979,0.0002795969,0.000319807,0.001059175,0.2878492,0.01100276,0.05912939,0.006831299,0.6214696],"study_design_scores_gemma":[0.00002359238,0.0001466732,0.001604087,0.00004315934,0.00008229415,0.0003694957,0.0001615081,0.9595872,0.009512382,0.02346365,0.004948393,0.00005748708],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.008858684,0.0002333005,0.9893049,0.000201734,0.00003707494,0.0001624382,0.00006745545,0.0006423616,0.0004920871],"genre_scores_gemma":[0.236706,0.0003677186,0.7599131,0.000194076,0.0001109215,0.0003118486,0.0004014562,0.0002319451,0.001763004],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.006211936,"threshold_uncertainty_score":0.03285223,"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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0003516751,0.0003433843,0.0001696201,0.0008288224,0.0004130742,0.0005018057,0.0004199656,0.0002772942,0.001121401],"category_scores_gemma":[0.001628519,0.0001050741,0.0002422959,0.0008692043,0.0003277355,0.0006715552,0.000172755,0.0003050108,0.0001685327],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0009293939,"about_ca_system_score_gemma":0.000393395,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.005453833,"about_ca_topic_score_gemma":0.003746123,"domain_scores_codex":[0.9996094,0.00003953495,0.00001402329,0.00006011602,0.0001420372,0.0001348652],"domain_scores_gemma":[0.9987699,0.0004503642,0.0002146323,0.0001082218,0.0003635653,0.00009322236],"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":[0.002127184,0.0006625974,0.1021006,0.0005443448,0.0002131114,0.00227638,0.0009427759,0.5808231,0.1905437,0.005882446,0.003944106,0.1099397],"study_design_scores_gemma":[0.00001960036,0.0006220329,0.0872904,0.00002236675,0.00006205483,0.0005645915,0.000573264,0.8664802,0.04094961,0.001118176,0.002254979,0.00004275012],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9959614,0.0001322498,0.00187663,0.00003328501,0.00001563302,0.00001315498,0.00009352032,0.0001744619,0.001699659],"genre_scores_gemma":[0.9988287,0.00003839056,0.0005920186,0.000007028732,0.000002442342,0.000005343588,0.0001193804,0.00002195198,0.0003845504],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.005453833,"threshold_uncertainty_score":0.01084417,"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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.001357003,0.0009013879,0.0008460685,0.005743628,0.0005287706,0.001279446,0.001008931,0.0007309441,0.002661395],"category_scores_gemma":[0.006069756,0.0002247502,0.0009673301,0.003030869,0.00035708,0.002144375,0.0009700509,0.0008967386,0.001140476],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.000529235,"about_ca_system_score_gemma":0.0009469751,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002128829,"about_ca_topic_score_gemma":0.003555547,"domain_scores_codex":[0.9991239,0.0002192402,0.0001062736,0.0002265137,0.000267908,0.0000562539],"domain_scores_gemma":[0.9982226,0.0008572129,0.0001229021,0.0003013319,0.000440715,0.00005515127],"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":[0.0001353956,0.00007786321,0.001802897,0.0003075593,0.000109113,0.0001106298,0.0001869682,0.01135264,0.00692225,0.006841643,0.005362442,0.9667906],"study_design_scores_gemma":[0.0000710713,0.0004389611,0.006090099,0.0001836923,0.0001848373,0.0008330988,0.0004857356,0.8659732,0.03199184,0.0320834,0.06154941,0.0001147769],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.01234555,0.0006683276,0.9788566,0.0001067033,0.00008069057,0.0001457302,0.0007415699,0.005443261,0.001611674],"genre_scores_gemma":[0.1115567,0.0005035115,0.8790146,0.00007612519,0.00008017879,0.0002196357,0.003896232,0.0004632237,0.004189839],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.005743628,"threshold_uncertainty_score":0.008903265,"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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.002233212,0.0009960689,0.0004886844,0.009624212,0.001628865,0.003308288,0.002554262,0.0008479102,0.01434087],"category_scores_gemma":[0.003504895,0.0004912663,0.001181979,0.01613973,0.001159496,0.001816935,0.001577013,0.000871616,0.004579602],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.007117357,"about_ca_system_score_gemma":0.01012005,"about_ca_topic_candidate":true,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.5319229,"about_ca_topic_score_gemma":0.5608607,"domain_scores_codex":[0.9988669,0.0001559219,0.00007845056,0.000123513,0.0006494956,0.0001257051],"domain_scores_gemma":[0.9980148,0.0003166341,0.00006830375,0.0001296997,0.001350084,0.0001204446],"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":[0.0000594965,0.00004362507,0.002524113,0.001070508,0.00006496831,0.0001702914,0.0006546628,0.01601214,0.001941166,0.03518027,0.07771523,0.8645636],"study_design_scores_gemma":[0.000007573211,0.00004966058,0.01080416,0.0009899486,0.00006763016,0.0003696972,0.001260383,0.04189201,0.003471927,0.02170607,0.919188,0.0001930393],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.01478672,0.04640619,0.6879966,0.006080321,0.001614953,0.0007143604,0.01562462,0.007566493,0.2192097],"genre_scores_gemma":[0.1091253,0.06923943,0.7330837,0.001200051,0.0006912067,0.0005728887,0.01587253,0.001461852,0.06875312],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.5319229,"threshold_uncertainty_score":0.9416673,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null}]}