> > My proposal was for: > > > отец<n><sg><gen><@subj><§agent><%:отца><:human><:kin><!:aef31><!:fcd32> > > If we have to have these "secondary tags"... which I have yet to be > completely convinced of, >
What exactly is your hesitation here? I want to make sure you guys are happy with the proposal before going ahead with it, but I'm not able to get through to you with arguments about eliminating trimming and markup handling. Given that there's no regression, and there's clear benefits, which were pointed out in the IRC, with regards to large monodixes and small bidixes creating a bottleneck which makes disambiguation and transfer demonstrably worse. Also, we discussed that after eliminating trimming we can weigh the analyses based on the bidix so as to keep the benefits of trimming with rare words and compounds. What is cause of hesitation to include secondary tags? > I would like to have them be readable and not clutter the stream with > unnecessary > verbosity. There are a lot of rule-based formalisms out there that are > impossible to read, > having been dreamt up by people who don't actually spend a lot of time > writing language > data, and I would like to avoid that happening with Apertium. > > Again, and again I want to see a translation and a linguistic > motivation. In an _actual_ > language pair, not in someone's imagination. I agree, and the contention here, more than just objective metrics, is also which will be more readable. Some might feel the prefixes interfere with the data, and some might feel they're self-documenting and clearer. Which is why this mail was sent, to find out the views of the people who work with actual language pairs and find out which is better for them. > We have a lot of modules that have been made but not reached use in a > released pair, > so I don't see how this should be different. > Forgive me if I misunderstood you but this is a little disheartening for me. I want to give this project my best, and I'm taking immense care to convince everyone of the benefits, of how there will be no regression, because I want to create a benefit for the released pairs. If this project is going to be shelved then there really isn't much use of me worrying so much about whether people see the linguistic motivation and benefits that come out of this. I've always seen myself as a linguistics student first and computer scientist second, and several times I have claimed that to eliminate trimming, which is the primary linguistic motivation, propagating surface form in the pipe is essential, for which secondary tags are necessary. I request the community to bombard this project with their skepticism, their doubts, their suggestions, their criticism, and I promise there will be thorough discussion and we will achieve an acceptable compromise on all fronts. What I would hate for this project is to be finished and then never reaching released pairs because some fundamental flaw was never discussed. Thanks and Regards, Tanmai Khanna -- *Khanna, Tanmai*
_______________________________________________ Apertium-stuff mailing list Apertium-stuff@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/apertium-stuff
