{"id":"W1997539970","doi":"10.1109/icassp.2010.5494926","title":"Large margin estimation of n-gram language models for speech recognition via linear programming","year":2010,"lang":"en","type":"article","venue":"","topic":"Speech Recognition and Synthesis","field":"Computer Science","cited_by":5,"is_retracted":false,"has_abstract":true,"ca_institutions":"York University","funders":"","keywords":"Discriminative model; Computer science; Margin (machine learning); Speech recognition; Language model; Viterbi algorithm; Vocabulary; Word error rate; Viterbi decoder; n-gram; Artificial intelligence; Pattern recognition (psychology); Decoding methods; Hidden Markov model; Algorithm; Machine learning","routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.001163395,0.001161225,0.0009237654,0.0003950848,0.0004102122,0.0007148699,0.001587122,0.0008998322,0.002872515],"category_scores_gemma":[0.005510891,0.0007371156,0.0005964655,0.0006033551,0.0006508305,0.001603945,0.001503226,0.002495227,0.002132675],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0005897288,"about_ca_system_score_gemma":0.0009051338,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00237606,"about_ca_topic_score_gemma":0.004271543,"domain_scores_codex":[0.9992238,0.0003419824,0.00003557613,0.0001885806,0.000159554,0.00005049106],"domain_scores_gemma":[0.998193,0.001249322,0.0001492404,0.0001824328,0.0001760256,0.00005008737],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0001971439,0.0001276439,0.000414752,0.0001475753,0.00006257185,0.0001128565,0.0001095612,0.503159,0.01920726,0.01464225,0.003525536,0.4582939],"study_design_scores_gemma":[0.000004701118,0.00001781522,0.00004996906,0.000003650051,0.000003301737,0.00001719342,0.000004344608,0.9933462,0.00219872,0.003897284,0.0004518793,0.000004971483],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.001702199,0.0000666393,0.9974576,0.00004556639,0.000007413834,0.000009313596,0.00001648578,0.0005092879,0.0001854854],"genre_scores_gemma":[0.1569889,0.0002538398,0.8382972,0.0001594183,0.00007715996,0.0002414469,0.0004989253,0.0004344679,0.003048601],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.002872515,"threshold_uncertainty_score":0.00960958,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02641845991397187,"score_gpt":0.2823456849418688,"score_spread":0.255927225027897,"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."}}