“1412-2”版本间的差异
来自cslt Wiki
（→2014-12-16） |
|||
| 第9行： | 第9行： | ||
* compare the different nlp tool ,like stanford-tool, ltp(hagongda-tool) ,fudan-tool,openNLP(not done) | * compare the different nlp tool ,like stanford-tool, ltp(hagongda-tool) ,fudan-tool,openNLP(not done) | ||
=2014-12-16= | =2014-12-16= | ||
| − | * fix the bug with caoli that mert doesn't match the QA | + | * fix the bug with caoli that mert doesn't match the QA(done) |
| − | * prepare the report tommorow. | + | * prepare the report tommorow.(not done) |
| + | =2014-12-16= | ||
| + | * find the mert bug that doesn't converge. | ||
| + | * prepare the report | ||
| + | * arrange the interview. | ||
| + | * code for query analysis. | ||
2014年12月17日 (三) 01:10的版本
2014-12-10
- find the bug that multi query in lucene and test(done)
- extract the entity recognition from xiaoI and add to framework(done)
- test the xiaoI spell check and add to framework.(done)
- compare the different nlp tool ,like stanford-tool, ltp(hagongda-tool) ,fudan-tool,openNLP(not done)
2014-12-11
- sort out the improved search method of lucene in wiki.（done）
- check the xiaoI framework aggin and find the usefully module . then check the deepQA in siri. (done)
- compare the different nlp tool ,like stanford-tool, ltp(hagongda-tool) ,fudan-tool,openNLP(not done)
2014-12-16
- fix the bug with caoli that mert doesn't match the QA(done)
- prepare the report tommorow.(not done)
2014-12-16
- find the mert bug that doesn't converge.
- prepare the report
- arrange the interview.
- code for query analysis.
