{"meta":{"query_hash":"d5e1e7be5d0e","filters":{"venue":"Proceedings of the Fifteenth ACM International Conference on Web Search and Data Mining"},"cohort_total":5,"direct_labels_cover":0,"predictions_cover":5,"exported":5,"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/d5e1e7be5d0e","api":"https://metacan.xera.ac/api/v1/cohort?venue=Proceedings+of+the+Fifteenth+ACM+International+Conference+on+Web+Search+and+Data+Mining"},"results":[{"id":"W4200631100","doi":"10.1145/3488560.3498498","title":"Differentially Private Ensemble Classifiers for Data Streams","year":2022,"lang":"en","type":"article","venue":"Proceedings of the Fifteenth ACM International Conference on Web Search and Data Mining","topic":"Data Stream Mining Techniques","field":"Computer Science","cited_by":4,"is_retracted":false,"has_abstract":true,"route_ca_aff":true,"route_ca_fund":true,"route_ca_venue":false,"route_about_ca":false,"ca_institutions":"Simon Fraser University","funders":"Natural Sciences and Engineering Research Council of Canada","keywords":"Data stream mining; Concept drift; Computer science; Competitor analysis; Data stream; Data mining; STREAMS; Private information retrieval; Black box; Ensemble forecasting; Machine learning; Regression; Artificial intelligence; Data modeling; Statistics; Database; Mathematics; Computer security","score_opus":0.16805348750242247,"score_gpt":0.35826392776550936,"score_spread":0.1902104402630869,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W4200631100","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":[0.016617276,0.00067647314,0.98032504,0.00045101586,0.00009239967,0.00005960232,0.00014652406,0.0008205526,0.0008111151],"genre_scores_gemma":[0.72229075,0.0009402772,0.26904267,0.0005449965,0.0006394147,0.00030066326,0.0010931373,0.00019224838,0.0049558915],"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","domain_scores_codex":[0.99576074,0.0012393912,0.00027135242,0.00089692924,0.0014521104,0.00037945085],"domain_scores_gemma":[0.9890175,0.005352222,0.0008302987,0.003001955,0.0014622634,0.00033575395],"candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.007954455,0.0011220024,0.0024518357,0.0012443194,0.0010238474,0.0020881651,0.0027568713,0.0020824743,0.0016108236],"category_scores_gemma":[0.019406775,0.0005956648,0.0010591061,0.0015917221,0.0010303564,0.0056590065,0.0030816637,0.0044933264,0.00077737635],"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":[0.000710766,0.0003223542,0.0041733026,0.000115619616,0.0002060609,0.00018494262,0.00023874138,0.5996289,0.0044890726,0.053987164,0.0071263765,0.32881677],"study_design_scores_gemma":[0.000011543385,0.000028669092,0.00012595263,0.000006115072,0.000012235318,0.000025926141,0.000009283705,0.9789596,0.0010359321,0.0190495,0.00072961627,0.0000057074617],"about_ca_topic_score_codex":0.0015092986,"about_ca_topic_score_gemma":0.0015755068,"teacher_disagreement_score":0.007954455,"about_ca_system_score_codex":0.0015490154,"about_ca_system_score_gemma":0.0015501516,"threshold_uncertainty_score":0.042067647},"labels":[],"label_agreement":null},{"id":"W4206865191","doi":"10.1145/3488560.3498419","title":"Modeling Scale-free Graphs with Hyperbolic Geometry for Knowledge-aware Recommendation","year":2022,"lang":"en","type":"article","venue":"Proceedings of the Fifteenth ACM International Conference on Web Search and Data Mining","topic":"Recommender Systems and Techniques","field":"Computer Science","cited_by":69,"is_retracted":false,"has_abstract":true,"route_ca_aff":true,"route_ca_fund":false,"route_ca_venue":false,"route_about_ca":false,"ca_institutions":"Huawei Technologies (Canada)","funders":"","keywords":"Knowledge graph; Computer science; Recommender system; Theoretical computer science; Semantics (computer science); Graph; Semantic space; ENCODE; Unification; Information retrieval; Artificial intelligence","score_opus":0.12534980454359115,"score_gpt":0.3362451936795552,"score_spread":0.21089538913596403,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W4206865191","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":[0.023097549,0.00053811475,0.973885,0.0002837816,0.000027180999,0.000027995502,0.0002500336,0.00053149165,0.0013587474],"genre_scores_gemma":[0.75420225,0.0013136476,0.23803273,0.00035525035,0.000099520315,0.00013281978,0.001230335,0.00015818489,0.00447524],"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","domain_scores_codex":[0.9993812,0.00018905418,0.00003065471,0.00020606889,0.00013549249,0.000057612477],"domain_scores_gemma":[0.99780375,0.0012419453,0.00026713125,0.00035030945,0.00025843194,0.00007835137],"candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0007923273,0.00087957527,0.00086870545,0.0014442962,0.00043209936,0.0012687516,0.0021395772,0.0013427141,0.001302483],"category_scores_gemma":[0.005151938,0.0008606961,0.0009905886,0.0017615829,0.00096491794,0.002914896,0.0011262606,0.0016417305,0.0006239363],"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":[0.00008172307,0.00005487773,0.0027625712,0.00013442154,0.00011586844,0.00012531198,0.00018431917,0.86759466,0.0024512212,0.048113815,0.002581221,0.075799994],"study_design_scores_gemma":[0.0000029556973,0.000008261488,0.00021306884,0.000004764241,0.000008576828,0.0000130543995,0.000007744492,0.9861459,0.00021399844,0.012905198,0.00047006286,0.0000064549495],"about_ca_topic_score_codex":0.022034558,"about_ca_topic_score_gemma":0.027285734,"teacher_disagreement_score":0.022034558,"about_ca_system_score_codex":0.0017525929,"about_ca_system_score_gemma":0.0008697614,"threshold_uncertainty_score":0.043812573},"labels":[],"label_agreement":null},{"id":"W4212816054","doi":"10.1145/3488560.3498376","title":"An Ensemble Model for Combating Label Noise","year":2022,"lang":"en","type":"article","venue":"Proceedings of the Fifteenth ACM International Conference on Web Search and Data Mining","topic":"Machine Learning and Data Classification","field":"Computer Science","cited_by":14,"is_retracted":false,"has_abstract":true,"route_ca_aff":true,"route_ca_fund":false,"route_ca_venue":false,"route_about_ca":false,"ca_institutions":"McMaster University","funders":"","keywords":"Computer science; Artificial intelligence; Noise (video); Machine learning; Artificial neural network; Consistency (knowledge bases); Pattern recognition (psychology); Image (mathematics); Set (abstract data type); Matching (statistics); Measure (data warehouse); Function (biology); Data mining; Mathematics; Statistics","score_opus":0.16405802616967322,"score_gpt":0.37567839057985014,"score_spread":0.21162036441017693,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W4212816054","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":[0.05571728,0.0005393912,0.94037473,0.0005009222,0.000093491275,0.000041065705,0.00015442252,0.0008699702,0.001708643],"genre_scores_gemma":[0.8532011,0.00047089506,0.13677795,0.0005031805,0.00027440907,0.00023137152,0.0007677271,0.00018478716,0.0075885914],"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","domain_scores_codex":[0.99901927,0.00025985463,0.00003952544,0.0003372939,0.00023253333,0.00011147871],"domain_scores_gemma":[0.99718034,0.0012897279,0.00029880775,0.00044776133,0.00066118286,0.00012210752],"candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0023202985,0.0013624694,0.001526968,0.00093187875,0.0006263201,0.0010538152,0.0027372793,0.0021755695,0.0014074758],"category_scores_gemma":[0.005789895,0.00062509073,0.0010513103,0.0008383285,0.00085073785,0.0035719802,0.0018613054,0.0027081205,0.0006104966],"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":[0.00025157034,0.00013814606,0.0026020934,0.00005184204,0.00010724866,0.000098689095,0.00014960289,0.8665044,0.0036119304,0.010362299,0.0025002917,0.11362188],"study_design_scores_gemma":[0.0000032866262,0.000021550883,0.00008137539,0.0000028147012,0.000010615695,0.000008193917,0.000003439185,0.99692684,0.0003087989,0.0024467485,0.0001832487,0.0000031085208],"about_ca_topic_score_codex":0.005502913,"about_ca_topic_score_gemma":0.007134654,"teacher_disagreement_score":0.005502913,"about_ca_system_score_codex":0.00089771656,"about_ca_system_score_gemma":0.0008551592,"threshold_uncertainty_score":0.012271106},"labels":[],"label_agreement":null},{"id":"W4213017138","doi":"10.1145/3488560.3498507","title":"A GNN-based Multi-task Learning Framework for Personalized Video Search","year":2022,"lang":"en","type":"article","venue":"Proceedings of the Fifteenth ACM International Conference on Web Search and Data Mining","topic":"Multimodal Machine Learning Applications","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":"Wilfrid Laurier University","funders":"","keywords":"Computer science; Task (project management); Human–computer interaction; Multimedia; Systems engineering; Engineering","score_opus":0.12830707427337426,"score_gpt":0.3845772540133766,"score_spread":0.2562701797400023,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W4213017138","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":[0.045623798,0.0017083599,0.94640285,0.0005752495,0.00013205981,0.00008369284,0.00037410556,0.0012850256,0.0038148435],"genre_scores_gemma":[0.84883,0.00072556705,0.14021793,0.0004808306,0.00015462682,0.00018105192,0.00076688937,0.0001123567,0.008530758],"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","domain_scores_codex":[0.9997081,0.00007314131,0.000014964453,0.00009822621,0.000050236264,0.000055140343],"domain_scores_gemma":[0.99966407,0.00014023234,0.000035937825,0.00002862531,0.0001038501,0.000027243519],"candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0006732659,0.0008031589,0.0009601598,0.00069061655,0.00031255832,0.0004717618,0.0015897835,0.0014049667,0.0017949116],"category_scores_gemma":[0.0017933454,0.0003969188,0.00070073846,0.00124641,0.00040998252,0.0012770188,0.0007206608,0.0012483122,0.00055394083],"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":[0.00013517393,0.00013012586,0.0012736518,0.0000958758,0.00006694506,0.00010307045,0.00007047534,0.8579677,0.0025356612,0.0059098983,0.0037329542,0.1279785],"study_design_scores_gemma":[0.0000033617102,0.000013296417,0.00008381304,0.0000018718981,0.000004471612,0.000008364587,0.0000029468642,0.998145,0.00009918498,0.0014849489,0.00015032648,0.0000023530417],"about_ca_topic_score_codex":0.027211271,"about_ca_topic_score_gemma":0.025866592,"teacher_disagreement_score":0.027211271,"about_ca_system_score_codex":0.0011530943,"about_ca_system_score_gemma":0.0009557578,"threshold_uncertainty_score":0.05410576},"labels":[],"label_agreement":null},{"id":"W4213235855","doi":"10.1145/3488560.3498387","title":"Translating Human Mobility Forecasting through Natural Language Generation","year":2022,"lang":"en","type":"article","venue":"Proceedings of the Fifteenth ACM International Conference on Web Search and Data Mining","topic":"Traffic Prediction and Management Techniques","field":"Engineering","cited_by":19,"is_retracted":false,"has_abstract":true,"route_ca_aff":true,"route_ca_fund":false,"route_ca_venue":false,"route_about_ca":false,"ca_institutions":"University of Waterloo","funders":"Australian Research Council","keywords":"Computer science; Bottleneck; Pipeline (software); Artificial intelligence; Natural language generation; Natural language; Machine learning; Mobility model; Data mining; Distributed computing","score_opus":0.13119351836417423,"score_gpt":0.33133583547336554,"score_spread":0.2001423171091913,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W4213235855","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":[0.037656553,0.0001961404,0.9483357,0.0007756231,0.00017813811,0.00023305224,0.0011305285,0.007975644,0.003518556],"genre_scores_gemma":[0.5065369,0.00038016666,0.4812418,0.00040876755,0.00016435975,0.00065818965,0.005566878,0.00044979528,0.004593191],"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","domain_scores_codex":[0.9994535,0.00015780932,0.00003206366,0.00021525017,0.00009753787,0.000043880616],"domain_scores_gemma":[0.9985103,0.00093459396,0.000076569675,0.00020186642,0.0002417592,0.00003487867],"candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.001033332,0.000937629,0.0005075831,0.0008493145,0.00038329142,0.000708042,0.0014516183,0.00082835264,0.0029939518],"category_scores_gemma":[0.0054686004,0.0004074717,0.0007472596,0.0008061808,0.000557036,0.001982463,0.0010421814,0.0011661769,0.0013419304],"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":[0.0002839039,0.00027567445,0.0035411974,0.00028225148,0.00005852364,0.0004733587,0.0005533082,0.49587968,0.01051188,0.022542294,0.016821107,0.44877687],"study_design_scores_gemma":[0.000014046122,0.000036071582,0.00022593663,0.000008978052,0.000009922817,0.000036082798,0.000042576627,0.9807912,0.002965896,0.01299145,0.0028664542,0.00001139972],"about_ca_topic_score_codex":0.0080671795,"about_ca_topic_score_gemma":0.0072780685,"teacher_disagreement_score":0.0080671795,"about_ca_system_score_codex":0.000845357,"about_ca_system_score_gemma":0.001246983,"threshold_uncertainty_score":0.016040444},"labels":[],"label_agreement":null}]}