Difference between revisions of "User:Fpetkovski"
Jump to navigation
Jump to search
Line 1: | Line 1: | ||
− | |||
==Reports== |
==Reports== |
||
* [[Lexical feature transfer - First report]] |
* [[Lexical feature transfer - First report]] |
||
Line 20: | Line 19: | ||
* '''Improve current method:''' |
* '''Improve current method:''' |
||
** Split test corpus in two (dev, test) |
** Split test corpus in two (dev, test) |
||
− | ** Rerun the experiments and check with test corpus |
+ | *** Rerun the experiments and check with test corpus |
− | ** Look at dev corpus to see what kind of patterns there are in lines that aren't getting matched |
+ | *** Look at dev corpus to see what kind of patterns there are in lines that aren't getting matched |
+ | ** Look at combining the 1-feature with the 2-feature model as backoff. |
||
Revision as of 12:35, 27 July 2012
Reports
TODO
- Try generating corpus from monolingual SL corpus:
- Оваа лабавост на регулативите се одразува врз третманот на уапсените корисници на дрога.
- Run through lexical transfer
-biltrans
- Run through
apertium-lex-tools/scripts/biltrans-to-multitrans.py
- Run through
apertium-lex-learner/irstlm-ranker
- Run through lexical transfer
- This will give:
- SL:TL selection possibilities
- probabilities from the TL language model for each selection
- Select a subset for training where one translation has a substantially higher proportion of the probability mass than the rest.
- Look at finding out how to work out what "substantially" should be.
- Оваа лабавост на регулативите се одразува врз третманот на уапсените корисници на дрога.
- Improve current method:
- Split test corpus in two (dev, test)
- Rerun the experiments and check with test corpus
- Look at dev corpus to see what kind of patterns there are in lines that aren't getting matched
- Look at combining the 1-feature with the 2-feature model as backoff.
- Split test corpus in two (dev, test)