{"title":"symgiz++:一个用于并行计算对称字对齐模型的工具","authors":"Marcin Junczys-Dowmunt, A. Szal","doi":"10.1109/IMCSIT.2010.5679866","DOIUrl":null,"url":null,"abstract":"SyMGiza++ - a tool that computes symmetric word alignment models with the capability to take advantage of multi-processor systems - is presented. A series of fairly simple modifications to the original IBM/Giza++ word alignment models allows to update the symmetrized models between each iteration of the original training algorithms. We achieve a relative alignment quality improvement of more than 17% compared to Giza++ and MGiza++ on the standard Canadian Hansards task, while maintaining the speed improvements provided by MGiza++'s capability of parallel computations.","PeriodicalId":147803,"journal":{"name":"Proceedings of the International Multiconference on Computer Science and Information Technology","volume":"49 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2010-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"5","resultStr":"{\"title\":\"SyMGiza++: A tool for parallel computation of symmetrized word alignment models\",\"authors\":\"Marcin Junczys-Dowmunt, A. Szal\",\"doi\":\"10.1109/IMCSIT.2010.5679866\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"SyMGiza++ - a tool that computes symmetric word alignment models with the capability to take advantage of multi-processor systems - is presented. A series of fairly simple modifications to the original IBM/Giza++ word alignment models allows to update the symmetrized models between each iteration of the original training algorithms. We achieve a relative alignment quality improvement of more than 17% compared to Giza++ and MGiza++ on the standard Canadian Hansards task, while maintaining the speed improvements provided by MGiza++'s capability of parallel computations.\",\"PeriodicalId\":147803,\"journal\":{\"name\":\"Proceedings of the International Multiconference on Computer Science and Information Technology\",\"volume\":\"49 1\",\"pages\":\"0\"},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2010-10-01\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"5\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Proceedings of the International Multiconference on Computer Science and Information Technology\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/10.1109/IMCSIT.2010.5679866\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Proceedings of the International Multiconference on Computer Science and Information Technology","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/IMCSIT.2010.5679866","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
SyMGiza++: A tool for parallel computation of symmetrized word alignment models
SyMGiza++ - a tool that computes symmetric word alignment models with the capability to take advantage of multi-processor systems - is presented. A series of fairly simple modifications to the original IBM/Giza++ word alignment models allows to update the symmetrized models between each iteration of the original training algorithms. We achieve a relative alignment quality improvement of more than 17% compared to Giza++ and MGiza++ on the standard Canadian Hansards task, while maintaining the speed improvements provided by MGiza++'s capability of parallel computations.