Volume 29, Issue 3 (9-2014)                   ... 2014, 29(3): 839-864 | Back to browse issues page

XML Persian Abstract Print


Download citation:
BibTeX | RIS | EndNote | Medlars | ProCite | Reference Manager | RefWorks
Send citation to:

Arabzadeh M, Araghi S M. Analysis of Azari Language based on Parsing using Link Gram. .... 2014; 29 (3) :839-864
URL: http://jipm.irandoc.ac.ir/article-1-2443-en.html
Islamic Azad University of Ahar
Abstract:   (4856 Views)
There are different classes of theories for the natural lanuguage syntactic parsing problem and for creating the related grammars .This paper presents a syntactic grammar developed in the link grammar formalism for Turkish which is an agglutinative language. In the link grammar formalism, the words of a sentence are linked with each other depending on their syntactic roles. Turkish has complex derivational and inflectional morphology, and derivational and inflection morphemes play important syntactic roles in the sentences. In order to develop a link grammar for Turkish, the lexical parts in the morphological representations of Turkish words are removed, and the links are created depending on the part of speech tags and inflectional morphemes in words. Furthermore, a derived word is separated at the derivational boundaries in order to treat each derivation morpheme as a special distinct word, and allow it to be linked with the rest of the sentence. The derivational morphemes of a word are also linked with each other with special links to indicate that they are parts of the same word. Finally the adapted unique link grammar formalism for Turkish provides flexibility for the linkage construction, and similar methods can be used for other languages with complex morphology. Finally, using the Delphi programming language, the link grammar related to the Azeri language was developed and implemented and then by selecting 250 random sentences, this grammar is evaluated and then tested. For 84.31% of the sentences, the result set of the parser contains the correct parse.
Full-Text [PDF 485 kb]   (1584 Downloads)    
Type of Study: Research | Subject: Library and Information Science
Received: 2014/03/9 | Accepted: 2014/03/9 | Published: 2014/03/9

Add your comments about this article : Your username or Email:
CAPTCHA

Send email to the article author


© 2020 All Rights Reserved | Iranian Journal of Information processing and Management

Designed & Developed by : Yektaweb