{"id":"W2397269111","doi":"","title":"Generalizing from Freebase and Patterns using Cluster-Based Distant Supervision for TAC KBP Slotfilling 2012.","year":2012,"lang":"en","type":"article","venue":"Theory and applications of categories","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":7,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"","keywords":"Computer science; Sentence; Classifier (UML); Word (group theory); Training set; Set (abstract data type); Natural language processing; Test set; Artificial intelligence; Information retrieval; Context (archaeology); Mathematics","routes":{"ca_aff":false,"ca_fund":false,"ca_venue":true,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.002224285,0.001201536,0.0009943992,0.001998916,0.0009188193,0.001344751,0.00246382,0.001528593,0.005391649],"category_scores_gemma":[0.01145527,0.0007353514,0.0008695766,0.002143955,0.0005112937,0.005274262,0.002667608,0.001989504,0.005811807],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0007684767,"about_ca_system_score_gemma":0.002136989,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01264438,"about_ca_topic_score_gemma":0.02174541,"domain_scores_codex":[0.9983332,0.000417985,0.000122451,0.0005414059,0.0004893882,0.00009562915],"domain_scores_gemma":[0.9958793,0.001726816,0.0001724678,0.00131435,0.0007795242,0.0001275494],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0005924228,0.0004923985,0.004279644,0.0009092119,0.0002290165,0.0005705504,0.001710011,0.02557194,0.03305556,0.005756414,0.1227911,0.8040417],"study_design_scores_gemma":[0.0001852163,0.0002260535,0.003811624,0.00009115722,0.0001303612,0.0006492995,0.0008374189,0.8477041,0.04358894,0.02935398,0.07330603,0.0001158119],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.07819864,0.0006637332,0.7704675,0.00107619,0.0002340981,0.0007466953,0.008318857,0.1326717,0.007622616],"genre_scores_gemma":[0.2955322,0.000250797,0.6513539,0.0004432971,0.0001099682,0.0005287014,0.039731,0.00381577,0.008234385],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.01264438,"threshold_uncertainty_score":0.02514154,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01750765833137411,"score_gpt":0.2808672439289689,"score_spread":0.2633595855975948,"validation_status":"score_only:v0-immature-baseline","note":"Baseline scores from an immature model (maturity gate not passed). Scores rank; they never assert a category."}}