{"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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.002688402,0.0009479278,0.001268961,0.01063249,0.001282757,0.00314184,0.001626944,0.001604866,0.01263392],"category_scores_gemma":[0.01055181,0.0001754927,0.001350508,0.00818453,0.0005278541,0.00227582,0.0009284728,0.001052098,0.006064916],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001499257,"about_ca_system_score_gemma":0.002048324,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.007187807,"about_ca_topic_score_gemma":0.005193616,"domain_scores_codex":[0.9960724,0.0004120425,0.0005268686,0.0008171496,0.001706147,0.0004653865],"domain_scores_gemma":[0.993431,0.001917231,0.0008484022,0.0009143848,0.002698848,0.0001901297],"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":[0.0003406648,0.0003852163,0.01226698,0.0005459996,0.0001319179,0.0002488313,0.00009840311,0.008859498,0.004293501,0.004572207,0.03025791,0.937999],"study_design_scores_gemma":[0.000125096,0.0008931126,0.05457021,0.0004001947,0.0002883496,0.00249836,0.0006411073,0.6990312,0.03246011,0.01649655,0.1923642,0.0002315441],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.2941181,0.01796066,0.49391,0.003969431,0.003033724,0.005114204,0.06452737,0.0146512,0.1027153],"genre_scores_gemma":[0.7021533,0.003335447,0.2294179,0.0003438088,0.0009665128,0.001241752,0.03314599,0.0002454686,0.02914982],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.01263392,"threshold_uncertainty_score":0.0422647,"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."}}