trochee: (Default)
trochee ([personal profile] trochee) wrote2004-10-04 12:36 pm

Reading list (back in school)

Reading updates for this week (now that classes have begun):

For Experimental Phonetics:
Acoustic and Auditory Phonetics, by Keith Johnson. So far, pretty easy summary of speech (audio) as sum of simple waves, plus some jargon review for digital signal processing.
I'm also rereading some old papers of mine from spring quarter, because I think I'll make my class project experiment be one that builds on some work I did for the same professor then.

For Computational Morphology:
Finite State Morphology, by Beesley and Karttunen. Also fairly easy so far; it's a review of the basics of transducers and regular expressions. So far it seems like a more-than-you-wanted-to-know user's manual for the xfst tool kit. Oh, and there's an embarrassing typo hidden on the cover ("Computational Linguisitics" [sic]).
An article "Principles of Semitic Word-Structure"; fairly challenging because it dives in assuming you know what's going on. It's also a fairly old-skool philology review, so it assumes you know or can gloss German, and tends to give unglossed examples in Arabic and Akkadian (!). But I think I've figured out what's going on.

[identity profile] elwe.livejournal.com 2004-10-04 10:10 pm (UTC)(link)
You're reading something on Semitic word structure for computational morphology? You don't happen to also be reading a paper by Stump and Finkel on Hebrew verb morphology with inheritence hierarchies, do you?

[identity profile] trochee.livejournal.com 2004-10-04 10:50 pm (UTC)(link)
no, but you should tell me more. please? or tell me where you found it so I can get a copy?

[identity profile] elwe.livejournal.com 2004-10-04 10:54 pm (UTC)(link)
You can find it at http://www.cs.um.edu.mt/~mros/WSL/papers/finkel:stump.pdf , I believe

It was a paper by one of my linguistics profs and one of my CS profs at UK; it's not really very practical, computationally, but it provides a nice language for describing languages' morphologies in a relatively human-friendly computational way