I think the point Andy was trying to make was why stop there? Why should EBMT be asked to compare to SMT but not rule-based MT? Why are SMT approaches not rejected for not comparing their results to state of the art EBMT and rule base MT? It would be great if every one had the resources to do this in every paper, but no one does. And given that no-one does, why are EBMT papers criticized for not making comparisons, while SMT papers aren't?
Hi Francis!
Seems to me ideally one would refer to all relevant work, up to some point of relevance. Not all SMT is relevant, but that which is -- pattern creation in general -- is. If *you* had to enumerate the work most relevant for EBMT (learning and matching and applying patterns for translation, which work would you use?).
E
-- Eduard Hovy email: [EMAIL PROTECTED] USC Information Sciences Institute tel: 310-448-8731 4676 Admiralty Way fax: 310-823-6714 Marina del Rey, CA 90292-6695 http://www.isi.edu/natural-language/nlp-at-isi.html
_______________________________________________ MT-List mailing list [EMAIL PROTECTED] http://www.computing.dcu.ie/mailman/listinfo/mt-list
