{"meta":{"query_hash":"7aa4c1c33ff6","filters":{"venue":"Journal of Natural Language Processing"},"cohort_total":4,"direct_labels_cover":0,"predictions_cover":4,"exported":4,"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/7aa4c1c33ff6","api":"https://metacan.xera.ac/api/v1/cohort?venue=Journal+of+Natural+Language+Processing"},"results":[{"id":"W2062246499","doi":"10.5715/jnlp.18.217","title":"Construction of Context Models for Word Sense Disambiguation","year":2011,"lang":"en","type":"article","venue":"Journal of Natural Language Processing","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":3,"is_retracted":false,"has_abstract":true,"route_ca_aff":true,"route_ca_fund":false,"route_ca_venue":false,"route_about_ca":false,"ca_institutions":"Université de Montréal","funders":"Ministry of Education, Culture, Sports, Science and Technology","keywords":"Word-sense disambiguation; Word (group theory); Context (archaeology); Computer science; SemEval; Natural language processing; Linguistics; Artificial intelligence; History; Philosophy; Engineering","score_opus":0.02621211769329632,"score_gpt":0.2813093315516963,"score_spread":0.25509721385839995,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W2062246499","genre_codex":"methods","genre_gemma":"methods","domain_codex":null,"domain_gemma":null,"model_version":"metacan-v3-hybrid-931329e0061c","genre_candidate":"methods","genre_consensus":"methods","domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.010118402,0.0008271072,0.9868897,0.00015150709,0.0000782435,0.000079279744,0.00020049694,0.0009084117,0.0007468697],"genre_scores_gemma":[0.27688676,0.0012692837,0.717652,0.00028654095,0.00021848867,0.0005450605,0.0013690097,0.00051456975,0.0012582691],"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","domain_scores_codex":[0.99774194,0.0008878513,0.00015347404,0.0007341198,0.0003645629,0.000118162716],"domain_scores_gemma":[0.9962,0.002347955,0.00028600116,0.0005329805,0.00047143686,0.00016156725],"candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0021081134,0.001839285,0.0017825026,0.004371137,0.0016630983,0.001797036,0.0017109954,0.0013379494,0.0017014336],"category_scores_gemma":[0.012522918,0.00147506,0.0022712606,0.0030186754,0.0010987713,0.005472577,0.002680236,0.0025213738,0.0010299447],"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":[0.00052952784,0.00029059636,0.006747219,0.00062370964,0.00066146714,0.00062687194,0.001250729,0.319558,0.013005146,0.115044855,0.008062129,0.5335999],"study_design_scores_gemma":[0.00003716352,0.000049518883,0.00065780943,0.00006449044,0.000094124414,0.00015254824,0.00010582369,0.8681103,0.0026168097,0.12255928,0.005503129,0.000049081398],"about_ca_topic_score_codex":0.00410116,"about_ca_topic_score_gemma":0.008894399,"teacher_disagreement_score":0.004371137,"about_ca_system_score_codex":0.0009183139,"about_ca_system_score_gemma":0.0017678065,"threshold_uncertainty_score":0.01114887},"labels":[],"label_agreement":null},{"id":"W2073971205","doi":"10.5715/jnlp.7.2_117","title":"The Exploration and Analysis of Using Multiple Thesaurus Types for Query Expansion in Information Retrieval.","year":2000,"lang":"en","type":"article","venue":"Journal of Natural Language Processing","topic":"Semantic Web and Ontologies","field":"Computer Science","cited_by":3,"is_retracted":false,"has_abstract":true,"route_ca_aff":false,"route_ca_fund":true,"route_ca_venue":false,"route_about_ca":false,"ca_institutions":"","funders":"Japan Society for the Promotion of Science; York University","keywords":"Thesaurus; Information retrieval; Query expansion; Computer science; Natural language processing","score_opus":0.017030922780763687,"score_gpt":0.28357993317831415,"score_spread":0.26654901039755047,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W2073971205","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":[0.15280297,0.010310872,0.8228758,0.0012792762,0.00014552091,0.00155311,0.00078477006,0.0017821436,0.008465605],"genre_scores_gemma":[0.2624173,0.0020035168,0.731969,0.00019511928,0.00007939637,0.0005589662,0.0008779292,0.00018968181,0.0017090781],"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","domain_scores_codex":[0.9921754,0.004461875,0.00078312895,0.0005837692,0.0018572609,0.00013850587],"domain_scores_gemma":[0.9799457,0.014270612,0.0011616511,0.0018906399,0.0024426298,0.00028870456],"candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.008368365,0.0006309839,0.00091470784,0.0065657236,0.000932851,0.0019497139,0.0013511131,0.0008385753,0.0014765742],"category_scores_gemma":[0.02976194,0.00067235285,0.0012870689,0.004403377,0.0009591386,0.00648105,0.001508527,0.00072122837,0.000678918],"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":[0.0009170882,0.0005097411,0.010139327,0.002651305,0.00052935525,0.00067822536,0.003083069,0.0109929275,0.051201504,0.023373246,0.005774187,0.89015007],"study_design_scores_gemma":[0.00044181728,0.0024083534,0.043113634,0.0019156504,0.0024188946,0.012789395,0.005792365,0.5754301,0.14883,0.076744914,0.12920792,0.00090695533],"about_ca_topic_score_codex":0.0023231956,"about_ca_topic_score_gemma":0.0026421405,"teacher_disagreement_score":0.008368365,"about_ca_system_score_codex":0.00082633714,"about_ca_system_score_gemma":0.0014966925,"threshold_uncertainty_score":0.044256628},"labels":[],"label_agreement":null},{"id":"W2316254469","doi":"10.5715/jnlp.17.2_51","title":"A Written Child Corpus with Editing History Tags","year":2010,"lang":"en","type":"article","venue":"Journal of Natural Language Processing","topic":"Natural Language Processing Techniques","field":"Computer Science","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":"Intecsea (Canada)","funders":"","keywords":"Computer science; Natural language processing; Linguistics; World Wide Web; Information retrieval; History; Philosophy","score_opus":0.004839098735990272,"score_gpt":0.23134153373138322,"score_spread":0.22650243499539294,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W2316254469","genre_codex":"other","genre_gemma":"dataset","domain_codex":null,"domain_gemma":null,"model_version":"metacan-v3-hybrid-931329e0061c","genre_candidate":"dataset","genre_consensus":null,"domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.28265312,0.002059133,0.08646283,0.003225854,0.0015264857,0.0011994814,0.2939183,0.012022782,0.31693205],"genre_scores_gemma":[0.6000476,0.00074769824,0.119245596,0.00042471918,0.00024464767,0.0014638873,0.20735264,0.0043797414,0.06609343],"study_design_codex":"design_other","study_design_gemma":"not_applicable","domain_scores_codex":[0.99837047,0.0007405399,0.00014182072,0.00038571976,0.00029135047,0.0000700693],"domain_scores_gemma":[0.9889326,0.008053843,0.00029187364,0.0012009498,0.0012598899,0.0002607941],"candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0012225541,0.00093647325,0.00048473562,0.0024269337,0.0016361612,0.0020261114,0.0011524541,0.0010831954,0.047262747],"category_scores_gemma":[0.008363669,0.0007373323,0.00038004338,0.0034073666,0.00138918,0.0018506558,0.0015063279,0.0018415475,0.009919518],"study_design_candidate":"not_applicable","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":[0.0019351337,0.00054314913,0.014957631,0.003889013,0.00013443665,0.010062837,0.045409102,0.00644786,0.032915514,0.06951451,0.39944443,0.41474643],"study_design_scores_gemma":[0.00028411506,0.0001824005,0.027009869,0.0005165873,0.0000959659,0.005261573,0.008233287,0.009245707,0.028404806,0.009970329,0.91062826,0.00016718931],"about_ca_topic_score_codex":0.023034979,"about_ca_topic_score_gemma":0.040778145,"teacher_disagreement_score":0.047262747,"about_ca_system_score_codex":0.0019422282,"about_ca_system_score_gemma":0.0026126613,"threshold_uncertainty_score":0.15810966},"labels":[],"label_agreement":null},{"id":"W3166208174","doi":"10.5715/jnlp.28.350","title":"The Effectiveness of Data Augmentation by Removing Unimportant sentence","year":2021,"lang":"en","type":"article","venue":"Journal of Natural Language Processing","topic":"Topic Modeling","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"route_ca_aff":false,"route_ca_fund":true,"route_ca_venue":false,"route_about_ca":false,"ca_institutions":"","funders":"Tokyo Metropolitan University; Institute for Catastrophic Loss Reduction","keywords":"Pointer (user interface); Computer science; Generator (circuit theory); Sentence; Natural language processing; Programming language; Artificial intelligence; Physics","score_opus":0.017118238281793714,"score_gpt":0.3082740527003838,"score_spread":0.2911558144185901,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W3166208174","genre_codex":"empirical","genre_gemma":"methods","domain_codex":null,"domain_gemma":null,"model_version":"metacan-v3-hybrid-931329e0061c","genre_candidate":"methods","genre_consensus":null,"domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.61033845,0.018444834,0.24226132,0.007138699,0.0073184846,0.0010808131,0.020915156,0.06416421,0.028338024],"genre_scores_gemma":[0.7193884,0.0020410055,0.234726,0.0017393441,0.000733923,0.0007416613,0.0275616,0.0012651493,0.01180285],"study_design_codex":"design_other","study_design_gemma":"bench_or_experimental","domain_scores_codex":[0.9964521,0.0014895998,0.00033672486,0.00094110327,0.0005348835,0.0002455381],"domain_scores_gemma":[0.98402184,0.009916619,0.00052281085,0.0031062877,0.0018480953,0.0005843085],"candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0049660583,0.003302474,0.0011553278,0.001663545,0.00091088016,0.0018030986,0.0017187995,0.0017227541,0.0051756063],"category_scores_gemma":[0.020553844,0.0007330351,0.001589163,0.00125575,0.0012437825,0.0044386154,0.0021082438,0.0029381951,0.004023359],"study_design_candidate":"bench_or_experimental","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":[0.00455403,0.00197714,0.0099315895,0.0013457624,0.00042696495,0.0004508857,0.00044274016,0.04812123,0.03591022,0.0026899185,0.06426332,0.82988626],"study_design_scores_gemma":[0.0008208987,0.0024709792,0.010508889,0.0003309722,0.00083370425,0.00074526615,0.0007974905,0.78720385,0.1472121,0.011766784,0.037073076,0.00023603896],"about_ca_topic_score_codex":0.0074228263,"about_ca_topic_score_gemma":0.007638136,"teacher_disagreement_score":0.0074228263,"about_ca_system_score_codex":0.0010100367,"about_ca_system_score_gemma":0.003402674,"threshold_uncertainty_score":0.026263356},"labels":[],"label_agreement":null}]}