Difference between revisions of "Turkish and Azerbaijani"

Revision as of 09:32, 25 April 2011

Fix punctuation -- 3271696
~~Proper names in caps~~
Subcategorise proper names as far as possible
Spaces in words in trmorph
subcategorise converb suffixes (See here)
subcategorise verbal noun suffixes (see here)
Check words in corpus which are analysed, but then have an apostrophe and unknown word after (e.g. case ending) ... perhaps they need to be added as proper nouns. e.g. Okyanusu'nun
find a way of exporting trmorph for apertium into two files, one with lexemes (that we specify -- e.g. the ones in the bilingual dictionary), and the other with the morphophonology.

Add consonant harmony |--| Azerbaijani, beside having double vowel harmony like turkish, has consonant harmony. Q/K (as well as their devoiced version ğ/y) change according to the precedent vowel.
Remove 040-exception_ben.fst |--| Unlike turkish, azerbaijani doesn't have irregular dative for men and sen (personal pronouns).
Disambiguation sucks big time: I don't know why it doesn't take içerler as <v><t_aor><3pp> but as a noun. Need to fix ASAP
Fix punctuation
Proper names in caps -- works in trmorph
Subcategorise proper names as far as possible
Interrogative mi
Add da/de cnjcoo as d<A>, since it follows vowel harmony
subcategorise converb suffixes (See here)
subcategorise verbal noun suffixes (see here)
remove apostrophe from case endings after propernames etc.

Finish adding closed categories
Fix punctuation
~~Numerals in the bidix from corpus~~
add proper support for compound numerals in the bidix
Finish adding nouns from the CSV list
try some sentences/paragraphs and fix disambiguation/transfer errors
write a script which generates stem lists for azmorph from the bilingual dictionary -- only stems which can be added automatically (some of them have e.g. <k> final or something)
continue adding words from the missing list
write up test cases on the wiki
find a girlfriend for the boss
the corpus is here, you'll need to clean it.
make a script to trim the trmorph lexicon to the bidix... we don't want to analyse anything more than we can translate !
FIX: yemek <v> does weird thing with the vowel

@@ Line 11: / Line 11: @@
 # subcategorise converb suffixes (See [https://sourceforge.net/mailarchive/forum.php?thread_name=20110422162330.GM2542%40rug.nl&forum_name=apertium-turkic here])
 # subcategorise verbal noun suffixes (see [https://sourceforge.net/mailarchive/forum.php?thread_name=20110422162330.GM2542%40rug.nl&forum_name=apertium-turkic here])
 # Check words in corpus which are analysed, but then have an apostrophe and unknown word after (e.g. case ending) ... perhaps they need to be added as proper nouns. e.g.  Okyanusu'nun
+# find a way of exporting trmorph for apertium into two files, one with lexemes (that we specify -- e.g. the ones in the bilingual dictionary), and the other with the morphophonology.
 ===Azmorph===