{"id":"W2119388680","doi":"10.3115/1220175.1220240","title":"Improved discriminative bilingual word alignment","year":2006,"lang":"en","type":"article","venue":"","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":72,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"","keywords":"Discriminative model; Generative grammar; Computer science; Word (group theory); Word error rate; Artificial intelligence; Machine translation; Natural language processing; Bilingual dictionary; Generative model; Selection (genetic algorithm); Translation (biology); Training set; Statistical model; Machine learning; Speech recognition; Linguistics","routes":{"ca_aff":false,"ca_fund":false,"ca_venue":false,"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":[0.001466411,0.001800226,0.00188914,0.001890951,0.001203729,0.001085319,0.001230443,0.0009354455,0.0087263],"category_scores_gemma":[0.004823803,0.0006926701,0.000937282,0.002915068,0.0005142392,0.00279272,0.00217125,0.001909866,0.01166325],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0006904691,"about_ca_system_score_gemma":0.002194911,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.006270446,"about_ca_topic_score_gemma":0.02126088,"domain_scores_codex":[0.9981104,0.0006357891,0.00009329621,0.0006857203,0.000299767,0.0001751043],"domain_scores_gemma":[0.9983071,0.0004541817,0.00009903803,0.0006050041,0.0004318899,0.0001028467],"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":[0.0006520112,0.0006921864,0.007996449,0.0004121463,0.0002107219,0.0004761292,0.0005208623,0.1232119,0.05382148,0.03079474,0.06569216,0.7155192],"study_design_scores_gemma":[0.0000940295,0.0001110817,0.001915374,0.00002458049,0.00007758222,0.0005375626,0.00009764326,0.9343882,0.02058539,0.02062165,0.02149252,0.00005444423],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.07741404,0.0009654794,0.8787384,0.0005840446,0.0004040095,0.0001279735,0.002684116,0.02355187,0.01553011],"genre_scores_gemma":[0.5195685,0.0004769563,0.4421212,0.0005952179,0.0002148005,0.0002051697,0.01780001,0.002812642,0.01620551],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.0087263,"threshold_uncertainty_score":0.02919239,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.00950746876754677,"score_gpt":0.2645663555949464,"score_spread":0.2550588868273996,"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."}}