{"id":"W2177050153","doi":"","title":"Feature Space Selection and Combination for Native Language Identification","year":2013,"lang":"en","type":"article","venue":"NPARC","topic":"Authorship Attribution and Profiling","field":"Computer Science","cited_by":14,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"","keywords":"Computer science; Classifier (UML); Artificial intelligence; Exploit; Feature selection; Spelling; Feature vector; Natural language processing; Support vector machine; Variety (cybernetics); Feature (linguistics); Machine learning; Linguistics","routes":{"ca_aff":false,"ca_fund":false,"ca_venue":true,"about_ca":true,"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":[0.005763473,0.00146414,0.001346126,0.003774508,0.0008497682,0.001828547,0.001101253,0.001091438,0.003461224],"category_scores_gemma":[0.01232501,0.0003667343,0.001310007,0.003040859,0.0004223837,0.00218201,0.001853211,0.001484738,0.003526892],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0003465574,"about_ca_system_score_gemma":0.0009804682,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001171214,"about_ca_topic_score_gemma":0.001675316,"domain_scores_codex":[0.9946619,0.002450141,0.0003543174,0.0006783088,0.001394895,0.0004605578],"domain_scores_gemma":[0.9938513,0.002124777,0.0002965456,0.001518029,0.001896087,0.0003132911],"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":[0.0006455235,0.0004373829,0.01096343,0.0001615592,0.0002194868,0.0001860766,0.0002108181,0.0150596,0.01543641,0.001541087,0.0182247,0.9369139],"study_design_scores_gemma":[0.0001139675,0.0009682165,0.01625651,0.00007384388,0.0002400509,0.00082827,0.0004805528,0.8786407,0.05857366,0.0162216,0.02739293,0.0002096515],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.2000184,0.002549661,0.7720265,0.00113343,0.000804671,0.0004150448,0.002261282,0.01515441,0.005636635],"genre_scores_gemma":[0.7606025,0.0003764198,0.2275987,0.0001720825,0.0002755443,0.0003728,0.00440181,0.0004163556,0.005783702],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.005763473,"threshold_uncertainty_score":0.0304805,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01409262470222885,"score_gpt":0.2742992865545968,"score_spread":0.260206661852368,"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."}}