{"id":"W2773837541","doi":"10.5430/air.v7n1p1","title":"Multiclass patent document classification","year":2017,"lang":"en","type":"article","venue":"Artificial Intelligence Research","topic":"Intellectual Property and Patents","field":"Business, Management and Accounting","cited_by":19,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"National Aeronautics and Space Administration","keywords":"Computer science; Support vector machine; Feature selection; Random forest; Artificial intelligence; C4.5 algorithm; Machine learning; Classifier (UML); Data mining; Document classification; Multiclass classification; Decision tree; Statistical classification; Naive Bayes classifier","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":"codex-gemma-dda1882f352a","candidate_categories":["sts","scholarly_communication","insufficient_payload"],"consensus_categories":["insufficient_payload"],"category_scores_codex":[0.002349005,0.0001510055,0.0001471698,0.000292763,0.002217609,0.002154103,0.001100183,0.0001024328,0.001355464],"category_scores_gemma":[0.001824294,0.0001234837,0.00008023897,0.0002644441,0.000450405,0.001311889,0.000517774,0.0004365175,0.01963033],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00008248022,"about_ca_system_score_gemma":0.00004064947,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.003696728,"about_ca_topic_score_gemma":0.0004617056,"domain_scores_codex":[0.9976704,0.00005108173,0.0003771096,0.0004416783,0.0007971661,0.0006625776],"domain_scores_gemma":[0.9980808,0.0001201334,0.0001519706,0.0008103884,0.0008029718,0.00003375379],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"theoretical_or_conceptual","study_design_scores_codex":[0.0003175542,0.0003515165,0.003568097,0.0001031918,0.00003498647,0.00003565235,0.0003885881,0.00007965017,0.01600819,0.4141131,0.00393061,0.5610689],"study_design_scores_gemma":[0.0001799938,0.0001538258,0.01121243,0.0002525526,0.00003653514,0.000003745925,0.004707242,0.2680464,0.08147896,0.4859517,0.1470175,0.0009591878],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.7821121,0.00008940222,0.006026572,0.01537223,0.002090429,0.001442428,0.000002940319,0.0002139097,0.19265],"genre_scores_gemma":[0.9969303,0.00004052671,0.0000836223,0.000236982,0.00118585,0.00006542117,0.00001003207,0.00002383025,0.001423443],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.5601097,"threshold_uncertainty_score":0.9995574,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.7736202196270415,"score_gpt":0.4283465409963569,"score_spread":0.3452736786306846,"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."}}