{"meta":{"page":1,"per_page":50,"max_per_page":100,"total":6,"total_is_capped":false,"direct_labels_cover":0,"predictions_cover":6,"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":"cffdf243d9d7","filters":{"venue":"Annals of Data Science"}},"results":[{"id":"W4289225573","doi":"10.1007/s40745-022-00432-6","title":"Forecasting Directional Movement of Stock Prices using Deep Learning","year":2022,"lang":"en","type":"article","venue":"Annals of Data Science","topic":"Stock Market Forecasting Methods","field":"Decision Sciences","cited_by":66,"is_retracted":false,"has_abstract":false,"routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false},"ca_institutions":"York University","funders":"","keywords":"Stock market; Computer science; Deep learning; Artificial intelligence; Stock market prediction; Word2vec; Machine learning; Stock (firearms); Sentiment analysis; Binary classification; Econometrics; Support vector machine; Economics; Engineering","authors":[{"name":"Deeksha Chandola","is_ca":true},{"name":"Akshit Mehta","is_ca":false},{"name":"Shikha Singh","is_ca":false},{"name":"Vinay Anand Tikkiwal","is_ca":false},{"name":"Himanshu Agrawal","is_ca":false}],"retraction":null,"screen_n_in":null,"score":{"opus":0.609124789376396,"gpt":0.5163442024371268,"spread":0.09278058693926927,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0004911183,0.0003495056,0.0003502946,0.0008273823,0.0001397724,0.0006572859,0.0004742944,0.0005137106,0.0009881576],"category_scores_gemma":[0.002608115,0.0003256925,0.0003566545,0.0007059944,0.0002266043,0.0007747129,0.0004167267,0.001004164,0.0002410615],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0004347275,"about_ca_system_score_gemma":0.0004257649,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.006671421,"about_ca_topic_score_gemma":0.01051164,"domain_scores_codex":[0.999913,0.00001497397,0.000005964443,0.00002376061,0.00002125468,0.00002107609],"domain_scores_gemma":[0.9992993,0.0003454615,0.0001309313,0.00005457325,0.0001189934,0.00005077486],"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":[0.0002723317,0.000236411,0.05923866,0.00007111341,0.0001589123,0.0001319885,0.00006555856,0.7617506,0.007790604,0.01011523,0.003038133,0.1571304],"study_design_scores_gemma":[0.000002678754,0.000004920785,0.001276619,0.000001899916,0.00000287964,0.000004012965,0.000003076938,0.9971693,0.0002256256,0.001241045,0.00006615098,0.000001806412],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.777115,0.0006170013,0.2186777,0.0005547665,0.0001721259,0.00002079777,0.0004182716,0.0003235759,0.002100727],"genre_scores_gemma":[0.9855433,0.0002028891,0.01246427,0.00003453763,0.00003629149,0.000007490637,0.0002875399,0.00001384097,0.001409988],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.006671421,"threshold_uncertainty_score":0.01326519,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null},{"id":"W4220667573","doi":"10.1007/s40745-022-00381-0","title":"A Machine Learning Model for Predicting Individual Substance Abuse with Associated Risk-Factors","year":2022,"lang":"en","type":"article","venue":"Annals of Data Science","topic":"Substance Abuse Treatment and Outcomes","field":"Medicine","cited_by":20,"is_retracted":false,"has_abstract":false,"routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false},"ca_institutions":"University of Waterloo","funders":"","keywords":"Machine learning; Decision tree; Naive Bayes classifier; Artificial intelligence; Random forest; Support vector machine; Substance abuse; Logistic regression; Computer science; CHAID; Vulnerability (computing); Psychology; Psychiatry; Computer security","authors":[{"name":"Uwaise Ibna Islam","is_ca":false},{"name":"Enamul Haque","is_ca":true},{"name":"Dheyaaldin Alsalman","is_ca":false},{"name":"Muhammad Nazrul Islam","is_ca":false},{"name":"Mohammad Ali Moni","is_ca":false},{"name":"Iqbal H. Sarker","is_ca":false}],"retraction":null,"screen_n_in":null,"score":{"opus":0.1783476059303608,"gpt":0.3609677976113045,"spread":0.1826201916809438,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.002995374,0.0007873368,0.001113205,0.001633692,0.0005483548,0.001215486,0.001319719,0.001184702,0.002514477],"category_scores_gemma":[0.009422171,0.0003276365,0.001202323,0.001224327,0.000328967,0.0009648955,0.000631653,0.001968492,0.0009902688],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0007371844,"about_ca_system_score_gemma":0.001394534,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01202877,"about_ca_topic_score_gemma":0.01050706,"domain_scores_codex":[0.999255,0.0003065939,0.00007267633,0.0002008538,0.00009080669,0.00007404298],"domain_scores_gemma":[0.9948384,0.004152909,0.0002463521,0.0002204953,0.0004169638,0.000124824],"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":[0.001117974,0.001567584,0.1753748,0.0001631537,0.00082974,0.0004216863,0.0002033251,0.4928887,0.001025542,0.006467648,0.01352716,0.3064127],"study_design_scores_gemma":[0.0000260755,0.00006480552,0.004220247,0.00001627146,0.00005379227,0.00006768639,0.00001595148,0.9908097,0.000125584,0.004180474,0.0004097735,0.000009716505],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.4399574,0.002438632,0.5379451,0.005776497,0.0005389114,0.0003033896,0.007901814,0.002024517,0.003113832],"genre_scores_gemma":[0.9315658,0.0005798295,0.05793056,0.0003453609,0.00031841,0.0002968298,0.004760377,0.00004121987,0.004161724],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.01202877,"threshold_uncertainty_score":0.02391756,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null},{"id":"W4295631129","doi":"10.1007/s40745-022-00446-0","title":"Impact of COVID-19 on Stock Indices Volatility: Long-Memory Persistence, Structural Breaks, or Both?","year":2022,"lang":"en","type":"article","venue":"Annals of Data Science","topic":"Market Dynamics and Volatility","field":"Economics, Econometrics and Finance","cited_by":6,"is_retracted":false,"has_abstract":false,"routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false},"ca_institutions":"Université Laval","funders":"","keywords":"Volatility (finance); Econometrics; Economics; Autoregressive conditional heteroskedasticity; Long memory; Financial economics; Markov chain; Portfolio; Stock market; Stock (firearms); Stochastic volatility; Coronavirus disease 2019 (COVID-19); Statistics; Mathematics; Geography","authors":[{"name":"Abdinardo Moreira Barreto de Oliveira","is_ca":false},{"name":"Anandadeep Mandal","is_ca":false},{"name":"Gabriel J. Power","is_ca":true}],"retraction":null,"screen_n_in":null,"score":{"opus":0.230385761088309,"gpt":0.3825578848856986,"spread":0.1521721237973896,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.01233243,0.0005791765,0.001293212,0.001191267,0.001326636,0.006155978,0.001806367,0.003012866,0.01809316],"category_scores_gemma":[0.0879911,0.000513597,0.001325621,0.001710092,0.001859248,0.005781033,0.002945488,0.005210213,0.002487191],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001288831,"about_ca_system_score_gemma":0.001733754,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0107457,"about_ca_topic_score_gemma":0.01016956,"domain_scores_codex":[0.9953791,0.001200059,0.000445753,0.001209926,0.0007026342,0.00106261],"domain_scores_gemma":[0.9333019,0.03313911,0.01136059,0.01317311,0.005224575,0.00380072],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"observational","study_design_scores_codex":[0.004395072,0.0008228355,0.7093176,0.000367038,0.001960069,0.001873404,0.0008762063,0.03780472,0.004428385,0.05614956,0.01927241,0.1627328],"study_design_scores_gemma":[0.000673457,0.001669533,0.6190487,0.0005388763,0.001996168,0.00137581,0.003518977,0.2483601,0.008687417,0.09088928,0.02295245,0.0002892017],"study_design_candidate":"observational","study_design_consensus":"observational","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9546021,0.003963625,0.009354516,0.01534353,0.0007924797,0.00003671519,0.002190981,0.0008125589,0.01290337],"genre_scores_gemma":[0.996231,0.0002831207,0.0004528521,0.0004923854,0.0001489726,0.000005479129,0.0007660675,0.00007949806,0.001540588],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.01809316,"threshold_uncertainty_score":0.06522095,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null},{"id":"W3031211888","doi":"10.1007/s40745-020-00291-z","title":"Empowering Data Mining Sciences by Habitual Domains Theory, Part II: Reaching Wonderful Solutions","year":2020,"lang":"en","type":"article","venue":"Annals of Data Science","topic":"AI-based Problem Solving and Planning","field":"Computer Science","cited_by":1,"is_retracted":false,"has_abstract":false,"routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false},"ca_institutions":"Carleton University","funders":"","keywords":"Management science; Computer science; Restructuring; Operations research; Decision analysis; Data science; Mathematics; Engineering; Political science; Mathematical economics","authors":[{"name":"Moussa Larbani","is_ca":true},{"name":"Po L. Yu","is_ca":false}],"retraction":null,"screen_n_in":null,"score":{"opus":0.3134005680289257,"gpt":0.3962351238778531,"spread":0.0828345558489274,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.00749786,0.0008695476,0.001206445,0.001941019,0.001235121,0.007221976,0.002494275,0.002192068,0.004135123],"category_scores_gemma":[0.0168427,0.0007872157,0.001513397,0.001938236,0.01098151,0.01840731,0.007656612,0.006355947,0.001051523],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001323667,"about_ca_system_score_gemma":0.003115944,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001032671,"about_ca_topic_score_gemma":0.001178438,"domain_scores_codex":[0.9962354,0.002255269,0.000206078,0.0004577362,0.0006661934,0.0001793331],"domain_scores_gemma":[0.9830184,0.01138598,0.0005427852,0.003518055,0.000856114,0.0006786332],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"theoretical_or_conceptual","study_design_scores_codex":[0.00002907681,0.00009994588,0.0009972599,0.0003310287,0.00006860596,0.00004510038,0.000615991,0.004459082,0.0005287091,0.9310476,0.004326753,0.05745104],"study_design_scores_gemma":[0.00000751245,0.00001037014,0.00009978338,0.00005438607,0.000006841996,0.00002102274,0.0001347633,0.009644287,0.000243027,0.9824427,0.007326624,0.000008474852],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":"theoretical_or_conceptual","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.01369212,0.006898822,0.9083322,0.04467675,0.0004956984,0.0001517374,0.0001945558,0.0005915295,0.0249665],"genre_scores_gemma":[0.3607705,0.01279017,0.6078675,0.007146494,0.001523829,0.0006640371,0.0004200213,0.0002531943,0.008564289],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.00749786,"threshold_uncertainty_score":0.03965294,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null},{"id":"W4401478708","doi":"10.1007/s40745-024-00570-z","title":"Non-negative Sparse Matrix Factorization for Soft Clustering of Territory Risk Analysis","year":2024,"lang":"en","type":"article","venue":"Annals of Data Science","topic":"Spatial and Panel Data Analysis","field":"Economics, Econometrics and Finance","cited_by":0,"is_retracted":false,"has_abstract":false,"routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false},"ca_institutions":"University of Guelph; Toronto Metropolitan University","funders":"","keywords":"Cluster analysis; Fuzzy logic; Data mining; Computer science; Fuzzy clustering; Sparse matrix; Matrix decomposition; Machine learning; Artificial intelligence","authors":[{"name":"Shengkun Xie","is_ca":true},{"name":"Chong Gan","is_ca":true},{"name":"Anna T. Ławniczak","is_ca":true}],"retraction":null,"screen_n_in":null,"score":{"opus":0.1723067282531261,"gpt":0.3603478806697574,"spread":0.1880411524166313,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.004556587,0.00165097,0.002903069,0.002222431,0.001240214,0.001972168,0.003368903,0.002518598,0.003603204],"category_scores_gemma":[0.0197429,0.001187128,0.002644585,0.002777014,0.002028881,0.002983491,0.002692453,0.003669793,0.001506283],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001542118,"about_ca_system_score_gemma":0.002338963,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01413317,"about_ca_topic_score_gemma":0.01834887,"domain_scores_codex":[0.9971685,0.00151498,0.0001417064,0.0005655393,0.0003647737,0.0002445094],"domain_scores_gemma":[0.9867524,0.009778928,0.0008328581,0.001240233,0.0009674134,0.0004282428],"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":[0.0003395232,0.0004319701,0.003364936,0.0005511662,0.0003961449,0.00021552,0.0004953096,0.6945394,0.002575293,0.08412452,0.01771495,0.1952513],"study_design_scores_gemma":[0.00001377887,0.00001695495,0.0001947863,0.00001519142,0.00001127676,0.00001835104,0.00003425657,0.9636115,0.0001191293,0.03548221,0.0004705071,0.0000120514],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.009356928,0.0004106257,0.98858,0.0003385268,0.0000508739,0.00007025434,0.0004017663,0.000367164,0.0004237404],"genre_scores_gemma":[0.3583144,0.0009120977,0.6299096,0.0005600968,0.0005074618,0.0006615076,0.004138443,0.0003401287,0.004656357],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.01413317,"threshold_uncertainty_score":0.0281018,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null},{"id":"W2937573193","doi":"10.1007/s40745-019-00216-5","title":"A New Family of Lifetime Distributions: Theory, Application and Characterizations","year":2019,"lang":"en","type":"article","venue":"Annals of Data Science","topic":"Statistical Distribution Estimation and Applications","field":"Mathematics","cited_by":0,"is_retracted":false,"has_abstract":false,"routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false},"ca_institutions":"University of Manitoba","funders":"","keywords":"Mathematics; Order statistic; Applied mathematics; Poisson distribution; Statistics; Estimator; Exponential family; Negative binomial distribution; Hazard","authors":[{"name":"Rasool Roozegar","is_ca":false},{"name":"G. G. Hamedani","is_ca":false},{"name":"Leila Amiri","is_ca":true},{"name":"Fatemeh Esfandiyari","is_ca":false}],"retraction":null,"screen_n_in":null,"score":{"opus":0.1660101840212524,"gpt":0.4371703354638459,"spread":0.2711601514425935,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.01368459,0.001795907,0.001991956,0.00723367,0.001922315,0.005992943,0.004337703,0.003215876,0.005923015],"category_scores_gemma":[0.05708092,0.00100725,0.002910945,0.005755641,0.004425969,0.01444507,0.004140554,0.006465808,0.001451665],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00274141,"about_ca_system_score_gemma":0.00211987,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001445491,"about_ca_topic_score_gemma":0.0008628925,"domain_scores_codex":[0.9951768,0.001530191,0.0003528056,0.001281749,0.001278853,0.0003795366],"domain_scores_gemma":[0.9597661,0.02336804,0.004510012,0.00449284,0.005976635,0.0018864],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"theoretical_or_conceptual","study_design_scores_codex":[0.00008047914,0.00006850724,0.005173387,0.0002185194,0.00008826792,0.0002324933,0.0005538031,0.01946732,0.001638743,0.9108688,0.006259754,0.05534998],"study_design_scores_gemma":[0.00003713735,0.0001062442,0.001827385,0.0001433248,0.00008114521,0.001693844,0.0002911499,0.2647123,0.0008558187,0.7129825,0.0171804,0.00008880683],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":"theoretical_or_conceptual","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.01239512,0.001923549,0.9809446,0.001372239,0.0001694385,0.00007127184,0.0004276315,0.0002292977,0.002466823],"genre_scores_gemma":[0.4798631,0.0177116,0.4644883,0.002288715,0.004100829,0.001116059,0.003106221,0.001035565,0.02628954],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.01368459,"threshold_uncertainty_score":0.0723719,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null}]}