Per Tunedal <[email protected]> čálii:
[...] > One big issue that I didn't solve was that the tagger had never been > trained - thus it made some very blatant errors. Further some of the new > vocabulary wouldn't show up in translations. Is the tagger trained now? Not by me at least, but the pair uses CG for pre-disambiguation now. I plan on writing more CG rules for Swedish→Norwegian, but I think any re-training of the tagger should wait until after that (otherwise you just have to train it yet again since major changes to CG would make the input not conform to the expectations of the tagger model). > An other problem was lexical selection. I'm eager to see how this is > handled now. See http://wiki.apertium.org/wiki/Rule-based_lexical_selection_module -Kevin
signature.asc
Description: PGP signature
------------------------------------------------------------------------------ Site24x7 APM Insight: Get Deep Visibility into Application Performance APM + Mobile APM + RUM: Monitor 3 App instances at just $35/Month Monitor end-to-end web transactions and take corrective actions now Troubleshoot faster and improve end-user experience. Signup Now! http://pubads.g.doubleclick.net/gampad/clk?id=272487151&iu=/4140
_______________________________________________ Apertium-stuff mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/apertium-stuff
