SMT Joshua
iteye_11686
这个作者很懒,什么都没留下…
展开
-
LongSentenceFilter Joshua SMT
I have been working on Joshua, a toolkit for SMT. Before extracting grammar from parallel corpus, one necessary step is to eliminate sentences of more than 100 words. For Hansard, it is common tha...2011-12-18 20:14:04 · 105 阅读 · 0 评论 -
LongSentenceFilter Joshua SMT [2]
Note that the first version of LongSentenceFilter is not complete, because even after filtering there still may be French sentences of more than 100 words. Now this version tackles this problem. Note ...原创 2012-02-11 05:52:54 · 126 阅读 · 0 评论