{"meta":{"query_hash":"cffdf243d9d7","filters":{"venue":"Annals of Data Science"},"cohort_total":6,"direct_labels_cover":0,"predictions_cover":6,"exported":6,"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/cffdf243d9d7","api":"https://metacan.xera.ac/api/v1/cohort?venue=Annals+of+Data+Science"},"results":[{"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,"route_ca_aff":true,"route_ca_fund":false,"route_ca_venue":false,"route_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","score_opus":0.1660101840212524,"score_gpt":0.43717033546384587,"score_spread":0.2711601514425935,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W2937573193","genre_codex":"methods","genre_gemma":"empirical","domain_codex":null,"domain_gemma":null,"model_version":"metacan-v3-hybrid-931329e0061c","genre_candidate":"empirical","genre_consensus":null,"domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.012395122,0.0019235491,0.98094463,0.0013722393,0.0001694385,0.00007127184,0.00042763146,0.00022929773,0.0024668232],"genre_scores_gemma":[0.4798631,0.017711604,0.4644883,0.0022887154,0.0041008294,0.0011160591,0.003106221,0.0010355649,0.026289538],"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"theoretical_or_conceptual","domain_scores_codex":[0.9951768,0.001530191,0.0003528056,0.0012817492,0.001278853,0.00037953656],"domain_scores_gemma":[0.9597661,0.023368042,0.004510012,0.0044928403,0.0059766355,0.0018863997],"candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.013684592,0.0017959072,0.001991956,0.0072336704,0.0019223145,0.005992943,0.004337703,0.003215876,0.0059230146],"category_scores_gemma":[0.057080917,0.0010072504,0.0029109449,0.0057556415,0.0044259694,0.014445069,0.0041405535,0.006465808,0.0014516652],"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":[0.00008047914,0.000068507245,0.0051733865,0.00021851936,0.000088267916,0.00023249334,0.00055380305,0.019467324,0.0016387429,0.9108688,0.006259754,0.055349976],"study_design_scores_gemma":[0.000037137346,0.00010624418,0.0018273848,0.00014332484,0.00008114521,0.0016938441,0.0002911499,0.26471227,0.00085581874,0.7129825,0.017180396,0.00008880683],"about_ca_topic_score_codex":0.0014454908,"about_ca_topic_score_gemma":0.00086289254,"teacher_disagreement_score":0.013684592,"about_ca_system_score_codex":0.00274141,"about_ca_system_score_gemma":0.0021198702,"threshold_uncertainty_score":0.0723719},"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,"route_ca_aff":true,"route_ca_fund":false,"route_ca_venue":false,"route_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","score_opus":0.31340056802892574,"score_gpt":0.39623512387785315,"score_spread":0.0828345558489274,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W3031211888","genre_codex":"methods","genre_gemma":"empirical","domain_codex":null,"domain_gemma":null,"model_version":"metacan-v3-hybrid-931329e0061c","genre_candidate":"empirical","genre_consensus":null,"domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.013692123,0.006898822,0.9083322,0.044676755,0.00049569836,0.00015173737,0.00019455576,0.00059152953,0.024966495],"genre_scores_gemma":[0.3607705,0.0127901705,0.60786754,0.0071464935,0.0015238294,0.0006640371,0.00042002133,0.0002531943,0.008564289],"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"theoretical_or_conceptual","domain_scores_codex":[0.99623543,0.0022552693,0.00020607795,0.00045773617,0.00066619343,0.0001793331],"domain_scores_gemma":[0.9830184,0.011385982,0.00054278516,0.003518055,0.00085611397,0.0006786332],"candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0074978597,0.0008695476,0.0012064448,0.0019410186,0.0012351207,0.007221976,0.0024942753,0.0021920677,0.0041351225],"category_scores_gemma":[0.016842704,0.0007872157,0.0015133965,0.0019382356,0.010981509,0.018407311,0.0076566124,0.0063559473,0.0010515227],"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":[0.000029076811,0.00009994588,0.0009972599,0.00033102866,0.000068605965,0.000045100376,0.00061599095,0.004459082,0.00052870915,0.93104756,0.004326753,0.057451036],"study_design_scores_gemma":[0.00000751245,0.000010370137,0.00009978338,0.000054386073,0.0000068419963,0.00002102274,0.00013476326,0.009644287,0.00024302704,0.98244274,0.007326624,0.000008474852],"about_ca_topic_score_codex":0.0010326707,"about_ca_topic_score_gemma":0.001178438,"teacher_disagreement_score":0.0074978597,"about_ca_system_score_codex":0.0013236665,"about_ca_system_score_gemma":0.0031159443,"threshold_uncertainty_score":0.039652944},"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,"route_ca_aff":true,"route_ca_fund":false,"route_ca_venue":false,"route_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","score_opus":0.17834760593036078,"score_gpt":0.36096779761130454,"score_spread":0.18262019168094376,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W4220667573","genre_codex":"methods","genre_gemma":"empirical","domain_codex":null,"domain_gemma":null,"model_version":"metacan-v3-hybrid-931329e0061c","genre_candidate":"empirical","genre_consensus":null,"domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.43995735,0.0024386316,0.5379451,0.005776497,0.0005389114,0.00030338956,0.007901814,0.0020245172,0.0031138323],"genre_scores_gemma":[0.93156576,0.0005798295,0.057930555,0.0003453609,0.00031840996,0.00029682976,0.0047603766,0.000041219868,0.004161724],"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","domain_scores_codex":[0.999255,0.00030659392,0.00007267633,0.00020085377,0.000090806694,0.000074042975],"domain_scores_gemma":[0.9948384,0.004152909,0.00024635214,0.00022049528,0.00041696383,0.00012482399],"candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.002995374,0.00078733684,0.0011132052,0.0016336923,0.0005483548,0.0012154862,0.0013197189,0.0011847019,0.002514477],"category_scores_gemma":[0.009422171,0.00032763646,0.0012023235,0.0012243267,0.00032896697,0.0009648955,0.000631653,0.001968492,0.0009902688],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_system_candidate":false,"about_ca_system_consensus":false,"study_design_scores_codex":[0.001117974,0.0015675842,0.17537478,0.00016315373,0.00082974,0.0004216863,0.00020332511,0.49288872,0.0010255418,0.006467648,0.013527161,0.30641267],"study_design_scores_gemma":[0.000026075502,0.00006480552,0.004220247,0.000016271464,0.00005379227,0.00006768639,0.000015951482,0.9908097,0.00012558399,0.004180474,0.00040977346,0.000009716505],"about_ca_topic_score_codex":0.01202877,"about_ca_topic_score_gemma":0.0105070565,"teacher_disagreement_score":0.01202877,"about_ca_system_score_codex":0.0007371844,"about_ca_system_score_gemma":0.0013945338,"threshold_uncertainty_score":0.023917556},"labels":[],"label_agreement":null},{"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,"route_ca_aff":true,"route_ca_fund":false,"route_ca_venue":false,"route_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","score_opus":0.609124789376396,"score_gpt":0.5163442024371268,"score_spread":0.09278058693926927,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W4289225573","genre_codex":"empirical","genre_gemma":"empirical","domain_codex":null,"domain_gemma":null,"model_version":"metacan-v3-hybrid-931329e0061c","genre_candidate":"empirical","genre_consensus":"empirical","domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.77711505,0.00061700126,0.21867774,0.0005547665,0.00017212593,0.000020797772,0.0004182716,0.00032357592,0.002100727],"genre_scores_gemma":[0.98554325,0.00020288906,0.012464267,0.000034537632,0.00003629149,0.000007490637,0.0002875399,0.000013840971,0.0014099876],"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","domain_scores_codex":[0.999913,0.000014973973,0.000005964443,0.000023760613,0.00002125468,0.000021076086],"domain_scores_gemma":[0.9992993,0.00034546154,0.00013093128,0.000054573247,0.000118993354,0.000050774855],"candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0004911183,0.00034950557,0.00035029463,0.0008273823,0.00013977235,0.00065728585,0.00047429441,0.0005137106,0.0009881576],"category_scores_gemma":[0.002608115,0.00032569253,0.00035665446,0.00070599443,0.00022660434,0.00077471294,0.00041672666,0.0010041638,0.00024106151],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_system_candidate":false,"about_ca_system_consensus":false,"study_design_scores_codex":[0.00027233167,0.00023641098,0.059238665,0.00007111341,0.00015891225,0.00013198848,0.00006555856,0.76175064,0.007790604,0.010115234,0.003038133,0.15713044],"study_design_scores_gemma":[0.0000026787543,0.0000049207847,0.0012766186,0.0000018999162,0.0000028796396,0.0000040129653,0.0000030769384,0.9971693,0.0002256256,0.001241045,0.00006615098,0.0000018064115],"about_ca_topic_score_codex":0.0066714208,"about_ca_topic_score_gemma":0.010511639,"teacher_disagreement_score":0.0066714208,"about_ca_system_score_codex":0.00043472747,"about_ca_system_score_gemma":0.00042576485,"threshold_uncertainty_score":0.0132651925},"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,"route_ca_aff":true,"route_ca_fund":false,"route_ca_venue":false,"route_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","score_opus":0.23038576108830897,"score_gpt":0.3825578848856986,"score_spread":0.15217212379738962,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W4295631129","genre_codex":"empirical","genre_gemma":"empirical","domain_codex":null,"domain_gemma":null,"model_version":"metacan-v3-hybrid-931329e0061c","genre_candidate":"empirical","genre_consensus":"empirical","domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.9546021,0.0039636246,0.009354516,0.015343533,0.00079247967,0.00003671519,0.0021909806,0.00081255886,0.012903366],"genre_scores_gemma":[0.996231,0.00028312067,0.00045285205,0.0004923854,0.0001489726,0.0000054791285,0.00076606753,0.00007949806,0.0015405876],"study_design_codex":"observational","study_design_gemma":"observational","domain_scores_codex":[0.9953791,0.0012000593,0.000445753,0.0012099259,0.00070263416,0.0010626096],"domain_scores_gemma":[0.9333019,0.033139106,0.011360594,0.013173105,0.0052245753,0.0038007204],"candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.012332435,0.0005791765,0.0012932122,0.0011912674,0.0013266363,0.0061559784,0.0018063671,0.003012866,0.018093163],"category_scores_gemma":[0.0879911,0.00051359704,0.0013256214,0.0017100924,0.001859248,0.005781033,0.0029454876,0.0052102134,0.0024871908],"study_design_candidate":"observational","study_design_consensus":"observational","about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_system_candidate":false,"about_ca_system_consensus":false,"study_design_scores_codex":[0.004395072,0.00082283554,0.70931756,0.00036703798,0.001960069,0.0018734036,0.0008762063,0.037804715,0.0044283853,0.056149557,0.019272413,0.1627328],"study_design_scores_gemma":[0.00067345705,0.0016695332,0.6190487,0.00053887634,0.0019961677,0.0013758099,0.0035189772,0.2483601,0.008687417,0.090889275,0.022952447,0.00028920174],"about_ca_topic_score_codex":0.010745701,"about_ca_topic_score_gemma":0.010169561,"teacher_disagreement_score":0.018093163,"about_ca_system_score_codex":0.0012888311,"about_ca_system_score_gemma":0.0017337541,"threshold_uncertainty_score":0.06522095},"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,"route_ca_aff":true,"route_ca_fund":false,"route_ca_venue":false,"route_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","score_opus":0.1723067282531261,"score_gpt":0.3603478806697574,"score_spread":0.18804115241663133,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W4401478708","genre_codex":"methods","genre_gemma":"empirical","domain_codex":null,"domain_gemma":null,"model_version":"metacan-v3-hybrid-931329e0061c","genre_candidate":"empirical","genre_consensus":null,"domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.009356928,0.00041062574,0.98858005,0.00033852676,0.000050873896,0.00007025434,0.00040176633,0.00036716397,0.0004237404],"genre_scores_gemma":[0.35831442,0.00091209775,0.6299096,0.0005600968,0.00050746184,0.0006615076,0.0041384427,0.00034012872,0.0046563568],"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","domain_scores_codex":[0.9971685,0.0015149802,0.00014170645,0.00056553935,0.00036477367,0.00024450937],"domain_scores_gemma":[0.9867524,0.009778928,0.0008328581,0.0012402333,0.0009674134,0.00042824278],"candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0045565874,0.0016509703,0.0029030694,0.002222431,0.0012402135,0.0019721678,0.0033689027,0.002518598,0.0036032042],"category_scores_gemma":[0.019742904,0.0011871278,0.0026445854,0.0027770137,0.002028881,0.0029834914,0.0026924529,0.0036697935,0.0015062834],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_system_candidate":false,"about_ca_system_consensus":false,"study_design_scores_codex":[0.00033952322,0.00043197014,0.0033649364,0.00055116625,0.0003961449,0.00021551998,0.0004953096,0.69453937,0.0025752934,0.08412452,0.017714953,0.19525127],"study_design_scores_gemma":[0.000013778874,0.000016954948,0.0001947863,0.00001519142,0.000011276759,0.000018351036,0.000034256573,0.9636115,0.00011912931,0.03548221,0.00047050708,0.000012051397],"about_ca_topic_score_codex":0.0141331665,"about_ca_topic_score_gemma":0.01834887,"teacher_disagreement_score":0.0141331665,"about_ca_system_score_codex":0.0015421185,"about_ca_system_score_gemma":0.0023389626,"threshold_uncertainty_score":0.028101802},"labels":[],"label_agreement":null}]}