mns meeting 27.6. 2023
Present: Csilla, Jack, Trond.
Agenda
-
Status
- Priorities onward
- Next meeting
Status
for the work
Coverage:
- Textbook: 230627: 1-381/3997 = 0.902
- Luima Seripos: 230627: 1-110000/713183 = 0.846
for the fst
Problems with verbs
Verbs and nouns are quite alike, but we look at different forms.
The verbal stem is close to the noun lemma. Trond has check
Priorities onward
Linguistic:
- Go through the missing lists of both LS and textbook.
- Look at LS missing lst and textbook missing list. Goal: Textbook 100% and LS 95%, thereafter LS better than 95%.
Summer
Csilla will work, and consult Trond or Jack whenever running into trouble.
working with missing lists
Two ways of working:
- Csilla adds missing words to fst and typos to typos.txt
- Csilla uses the file.missing.freq.date file and writes comments to each word, one of the following:
- fixed (or: no comment means “Csilla fixed it”)
- correct # this means: must be fixed in fst
- typo > correctform # means: Csilla adds it to typos.txt
Trond to make sure the missing lists are run without long vowel problems.
Working with Yamls
Csilla to make sure they are ok. Lemmas should be tested only once.
Next meeting
We will look for a date it in a month or so (perhaps less).