Some peers argue that rich information based grammar is more important than parsing technique for these languages. How will you take this argument?
> How will you take this argument?
I take it as too terse.
But a very good parsing technique will not help you much with a grammar that is a poor description of the language.
And more generally, you should not expect too much from algorithms that are fed with inadequate data.
GIGO
Encoding of morphological information at POS, syntactic and functional level of annotation has higher precedence in case of MRLs.
Along with the parsing technique, the more important thing for parsing a MRL with free order is the sufficient encoding of morphological information at POS and syntactic level. This has been experienced by myself and can also seen in Tsarfaty's articles on MRLs.
- Badges
- Science topic
- Similar topics
- Phytochemistry
- Extracts