The Khasi shallow parser identifies noun phrases and verb phrases only. The parser is the first of its kind for the language, where it was developed using the Hidden Markov Model (HMM) which utilizes a training corpus comprising of 24,194 chunks of noun and verb chunks out of a total of 3,997 sentences and 86,087 tokens. The approach in developing the parser is taken as a tagging problem. Shallow parsing is a process where a text is divided into non-recursive syntactical units such as noun phrases, verb phrases, etc. Non-recursive implies these phrases or chunks are non-overlapping and do not contain each other as proposed by Abney (1991). Basically, the parser will identify a noun phrase or a verb phrase present in a sentence where NP indicates the starting of a noun phrase and VP indicates the staring of a verb phrase. The details of what constitute a Khasi noun phrase or verb phrase and the details of the parser construction are given in the paper titled Khasi Shallow Parser and others.
Katto katne shaphang ka parser
Ka parser ruh ka shim ïa ka senten ha ka ktien Khasi bad ka pyni kino ki kynhun kyntien kiba dei ki noun phrase (NP) bad kino ki kynhun kyntien kiba dei ki verb phrase (VP). Kane ka software ka pyni tang ki NP ne ki VP kiba ïeng marwei marwei, bad ka jingbatai bniah shaphang kane ka parser ka don ha ka jingthoh Khasi Shallow Parser. Sa shisien, kane ka dei ka parser kaba nyngkong kaba don ha ka ktien Khasi.
We use cookies to analyze website traffic and optimize your website experience. By accepting our use of cookies, your data will be aggregated with all other user data.