140428-Ma xi

来自cslt Wiki
2014年4月28日 (一) 11:31Mx讨论 | 贡献的版本

(差异) ←上一版本 | 最后版本 (差异) | 下一版本→ (差异)
跳转至: 导航搜索

Last week:

1.Learning the knowledge of language model

2.Segment the training/test sentences with the 150k lexicon; Get the ppl of the

test: /nfs/home/zhangzhiyong/work/train_470h/test/huawei_disanpi.txt. Using the following LM:

/home/thdnn/resource/lm/Hunhe_zhongzi_and_add_and_PPL_5yuan_1e9.lm

3.Build the new LM using the lexicon with the keywords involved; Re-segment the test files, and test the PPL.

This week:

To extract sentences of the related field from the original corpus.