<spectie> no obvious POS, but i suppose you can tell in some cases
<spectie> from the verb morphology in kk and in english
<nathan0n5ire> the words that end in - are verbs
<spectie> and sometimes it has POS
<spectie> АЙЛЫ /adj./ moonlight, moonlit.
<nathan0n5ire> yes adjectives are marked
<nathan0n5ire> I don't think the other ones are though
<spectie> it seems like kazakh stems are marked by being only in uppercase
<spectie> and also in cyrillic (obv)
<spectie> so something could be done with that
<spectie> to extract just the stems
<nathan0n5ire> all of the kazakh words are in uppercase
<spectie> АЙЛЫҚ monthly; monthly wages; ... MEP-31M period of one month.
<spectie>
<spectie> ah
<nathan0n5ire> sometimes the ocr also messed up
<nathan0n5ire> like AJIMAJIA- to embrace.
<spectie> MER-31M is ocr error
<spectie> i would approach this in passes
<spectie> i would start by extracting the really easy stuff
<spectie> where you just have two words:
<spectie> АЙТУ pronunciation.
<spectie> АЙМАҚТЫҚ regional.
<spectie> АЙНАЛАДА around.
<spectie>
<spectie> etc.
<spectie> i would put these in a separate file
<spectie> then i would extract the ones with only comma + full stop as punctuation
<spectie> АЙЛАКЕР sly, cunning one.
<spectie> АЙЛАКЕРЛІК slyness, cunning.
<spectie> АЙЛАЛЫ adroit, resourceful.
<spectie> АЙЛАСЫЗ artless, unsophisticated.
<spectie> etc.
<nathan0n5ire> around 8K start with a kazakh character
<nathan0n5ire> *8K lines
<spectie> nice
<spectie> lines starting with a kazakh character ?
<nathan0n5ire> [АаӘәБбВвГгҒғДдЕеЁёЖжЗзИиЙйКкҚқЛлМмНнҢңОоӨөПпРрСсТтУуҰұҮүФфХхҺһЦцЧчШшЩщЪъЫыІіЬьЭэЮюЯя]