Fabian, Exactly! This assumption is at the core of my research. Mathieu, I'd be interested in your analysis. If you're interested, we're open to collaborating to extend the test cases to the kinds of annotations that you need.
Cheers Pablo On Mar 15, 2012 8:10 AM, "Fabian Christ" <[email protected]> wrote: > > Hi, > > it would be great to have auch an analysis, but it is a common problem with > NLP techniques. The results are dependent on the used language, domain and > use case. It is impossible to write an enhancer that fits in all cases. > > So having a classification schema to specify what an enhancer is optimized > for and which assumptions are made, may be a way to improve the > documentation. > > - Fabian > Am 15.03.2012 07:49 schrieb "Mathieu D'Aquin" <[email protected]>:
