140519 - Xi Ma
From cslt Wiki
Revision as of 04:55, 19 May 2014 by Mx
1.Extract sentences in different ppl range from baiduzhidao corpus, which is merged with 28M corpus to construct the new training set.
Train a new language using this training set and test the ppl. Compare the result with xiaxi's method.
2.Finish the new vovabulary and mark the pinyin.