{"id":"W2166202273","doi":"","title":"Bootstrapping a Stochastic Transducer for Arabic-English Transliteration Extraction","year":2007,"lang":"en","type":"article","venue":"Meeting of the Association for Computational Linguistics","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":24,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Alberta","funders":"","keywords":"Bootstrapping (finance); Computer science; Transducer; Scripting language; Artificial intelligence; Natural language processing; Metric (unit); Transliteration; Task (project management); Speech recognition; Similarity (geometry); Arabic; Pattern recognition (psychology); Programming language; Linguistics; Acoustics; Engineering; Mathematics","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":[0.001306289,0.0006978076,0.00083025,0.0006630222,0.0004962951,0.000598632,0.00113502,0.001029932,0.002647689],"category_scores_gemma":[0.006989904,0.0004209183,0.0006879524,0.0005384317,0.0005704576,0.001294449,0.001257257,0.001568389,0.002431586],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0003981024,"about_ca_system_score_gemma":0.001077207,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002232038,"about_ca_topic_score_gemma":0.003744778,"domain_scores_codex":[0.9989686,0.0003767161,0.00006543864,0.000287454,0.0002142595,0.00008745825],"domain_scores_gemma":[0.996691,0.002254731,0.0001374763,0.0002654483,0.0005512852,0.00009999697],"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":[0.0005887948,0.0002849486,0.00450803,0.0002080847,0.0001122045,0.0004998581,0.0005159595,0.1227352,0.1010199,0.008974833,0.006438963,0.7541133],"study_design_scores_gemma":[0.00001138652,0.00009159778,0.0005117472,0.00000664707,0.00001456573,0.0001096164,0.00003794138,0.9742848,0.01938737,0.004316766,0.001210439,0.00001699363],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.04707269,0.0001076748,0.9468288,0.0003035283,0.00006859448,0.0000640031,0.000211931,0.004396667,0.0009460948],"genre_scores_gemma":[0.6380522,0.0001186249,0.3563839,0.0003169891,0.00009324498,0.000271262,0.001546176,0.0003394151,0.002878192],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.002647689,"threshold_uncertainty_score":0.008857429,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01409818308605324,"score_gpt":0.2948093699806726,"score_spread":0.2807111868946193,"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."}}