{"meta":{"query_hash":"65a47c5cc863","filters":{"venue":"2021 International Conference on Data Mining Workshops (ICDMW)"},"cohort_total":3,"direct_labels_cover":0,"predictions_cover":3,"exported":3,"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/65a47c5cc863","api":"https://metacan.xera.ac/api/v1/cohort?venue=2021+International+Conference+on+Data+Mining+Workshops+%28ICDMW%29"},"results":[{"id":"W3203427563","doi":"10.1109/icdmw53433.2021.00085","title":"Determining Standard Occupational Classification Codes from Job Descriptions in Immigration Petitions","year":2021,"lang":"en","type":"article","venue":"2021 International Conference on Data Mining Workshops (ICDMW)","topic":"Artificial Intelligence in Law","field":"Social Sciences","cited_by":1,"is_retracted":false,"has_abstract":true,"route_ca_aff":true,"route_ca_fund":false,"route_ca_venue":false,"route_about_ca":false,"ca_institutions":"Becton Dickinson (Canada)","funders":"","keywords":"Computer science; Code (set theory); Task (project management); Variety (cybernetics); Quality (philosophy); Natural language processing; Machine learning; Artificial intelligence; Programming language; Engineering","score_opus":0.31506050833024407,"score_gpt":0.43601401014173113,"score_spread":0.12095350181148706,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W3203427563","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":[0.8267258,0.00025761477,0.1107918,0.00091444625,0.00020032239,0.00065908796,0.03918632,0.0021066822,0.0191579],"genre_scores_gemma":[0.8324961,0.00018237896,0.09072665,0.00013605664,0.000031768974,0.00048081265,0.07179544,0.00019624115,0.0039546043],"study_design_codex":"observational","study_design_gemma":"simulation_or_modeling","domain_scores_codex":[0.9977174,0.00071960845,0.00027963394,0.00030164648,0.0008128718,0.00016898513],"domain_scores_gemma":[0.9862684,0.0069958805,0.0018170956,0.0015438221,0.0031073517,0.0002675308],"candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0031127157,0.00039115272,0.00032157908,0.00449339,0.0006983392,0.0014799922,0.000930036,0.00062426645,0.002514808],"category_scores_gemma":[0.021416824,0.00029512533,0.0004788305,0.003437024,0.0005162767,0.0012650737,0.0011793228,0.0009500277,0.0018684914],"study_design_candidate":"simulation_or_modeling","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":[0.0006577399,0.0007918044,0.5112777,0.00045266972,0.000078031626,0.0006922634,0.0023721477,0.08212549,0.003625674,0.016108884,0.06096434,0.3208533],"study_design_scores_gemma":[0.00006690423,0.00014933043,0.18422285,0.0003775112,0.00003396914,0.00030078203,0.0038535532,0.7381531,0.009218329,0.01744028,0.046081588,0.0001017848],"about_ca_topic_score_codex":0.029102778,"about_ca_topic_score_gemma":0.038282655,"teacher_disagreement_score":0.029102778,"about_ca_system_score_codex":0.0016144502,"about_ca_system_score_gemma":0.0024453204,"threshold_uncertainty_score":0.057866752},"labels":[],"label_agreement":null},{"id":"W4205504967","doi":"10.1109/icdmw53433.2021.00084","title":"Detection of Similar Legal Cases on Personal Injury","year":2021,"lang":"en","type":"article","venue":"2021 International Conference on Data Mining Workshops (ICDMW)","topic":"Artificial Intelligence in Law","field":"Social Sciences","cited_by":6,"is_retracted":false,"has_abstract":true,"route_ca_aff":true,"route_ca_fund":false,"route_ca_venue":false,"route_about_ca":true,"ca_institutions":"Queen's University","funders":"","keywords":"Personal injury; Computer science; Plaintiff; Compensation (psychology); Similarity (geometry); Artificial intelligence; Deep learning; Variation (astronomy); Legal research; Feature (linguistics); Style (visual arts); Data science; Natural language processing; Information retrieval; Psychology; Law; Social psychology; Political science; Linguistics","score_opus":0.2099168421139607,"score_gpt":0.4157965502019689,"score_spread":0.20587970808800818,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W4205504967","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":[0.93757015,0.0008676875,0.050291456,0.0013744379,0.00010641318,0.00058034394,0.0025424655,0.0010411406,0.005625943],"genre_scores_gemma":[0.95428705,0.0002141003,0.038809232,0.0002031727,0.000046416197,0.00009328869,0.0043869363,0.000026666403,0.001933065],"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","domain_scores_codex":[0.99727553,0.00048213842,0.00027839997,0.0006624139,0.0009803615,0.00032121997],"domain_scores_gemma":[0.9932106,0.0034036865,0.0010575952,0.0006134824,0.0013536578,0.00036104428],"candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0024099457,0.00045524864,0.00047578625,0.0074631507,0.0012155679,0.0013362551,0.0014595899,0.0019245525,0.0015223119],"category_scores_gemma":[0.013491231,0.00020480761,0.00067905773,0.002009531,0.001074162,0.0014564396,0.001519593,0.00093491445,0.0005689394],"study_design_candidate":"simulation_or_modeling","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":[0.0010478048,0.0014433088,0.3432822,0.0008284376,0.00025870118,0.014240709,0.003801829,0.069203496,0.020786896,0.010405985,0.03260595,0.50209475],"study_design_scores_gemma":[0.00006362109,0.00022221745,0.10176282,0.00020364438,0.00010890156,0.0035097583,0.0032611934,0.8395864,0.024638196,0.011620478,0.014953254,0.00006955922],"about_ca_topic_score_codex":0.020743275,"about_ca_topic_score_gemma":0.037842568,"teacher_disagreement_score":0.020743275,"about_ca_system_score_codex":0.0017443516,"about_ca_system_score_gemma":0.0016822065,"threshold_uncertainty_score":0.041245103},"labels":[],"label_agreement":null},{"id":"W4205619838","doi":"10.1109/icdmw53433.2021.00020","title":"Sentiment Analysis Using Part-of-Speech-Based Feature Extraction and Game-Theoretic Rough Sets","year":2021,"lang":"en","type":"article","venue":"2021 International Conference on Data Mining Workshops (ICDMW)","topic":"Sentiment Analysis and Opinion Mining","field":"Computer Science","cited_by":9,"is_retracted":false,"has_abstract":true,"route_ca_aff":true,"route_ca_fund":true,"route_ca_venue":false,"route_about_ca":false,"ca_institutions":"University of Regina","funders":"Natural Sciences and Engineering Research Council of Canada; University of Regina","keywords":"Computer science; Sentiment analysis; Artificial intelligence; Naive Bayes classifier; Rough set; Support vector machine; Ambiguity; Decision tree; Probabilistic logic; Feature extraction; Natural language processing; Data mining; Machine learning","score_opus":0.09865835572894807,"score_gpt":0.3584949497619248,"score_spread":0.25983659403297676,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W4205619838","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":[0.043385662,0.00037261273,0.9533162,0.0004996005,0.00007804884,0.00019452989,0.00030190294,0.00037184774,0.001479465],"genre_scores_gemma":[0.71593356,0.0004349501,0.28095743,0.00019326128,0.00012923157,0.0003627418,0.00074155245,0.00003508661,0.0012121978],"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","domain_scores_codex":[0.99792933,0.00080186425,0.00021000973,0.00032530728,0.0006085266,0.00012501184],"domain_scores_gemma":[0.9974486,0.0015310891,0.0003439496,0.00013982525,0.00047898781,0.000057497014],"candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0026262589,0.0010283701,0.0015135068,0.0023073754,0.00047730756,0.001569815,0.0009987896,0.0006487104,0.00073578244],"category_scores_gemma":[0.007257732,0.0003871802,0.0022873473,0.001273389,0.0005342586,0.0020626974,0.0006671609,0.0012048348,0.00030868396],"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":[0.0006010846,0.00043015505,0.010859789,0.00057394354,0.00072534167,0.00050091627,0.0006985149,0.50519496,0.015848935,0.036840748,0.006657256,0.42106846],"study_design_scores_gemma":[0.000011564338,0.000054443175,0.0010312904,0.000010773803,0.000039729654,0.000034612098,0.00003092511,0.9881458,0.0011744036,0.008965041,0.00048278464,0.000018635339],"about_ca_topic_score_codex":0.0035644225,"about_ca_topic_score_gemma":0.0023704672,"teacher_disagreement_score":0.0035644225,"about_ca_system_score_codex":0.0011092674,"about_ca_system_score_gemma":0.00081140845,"threshold_uncertainty_score":0.013889134},"labels":[],"label_agreement":null}]}