header
Image from OpenLibrary

Semantic extraction of Arabic multi-word expressions / Samah Meghawry Mohamed Elsayed ; Supervised Akram Ibrahim Salah , Abeer Elkorany , Tarek Elghazaly

By: Contributor(s): Material type: TextTextLanguage: English Publication details: Cairo : Samah Meghawry Mohamed Elsayed , 2016Description: 94 P. : charts , facsimiles ; 30cmOther title:
  • الاستخراج الدالالي للمصطلحات متعددة الكلمات باللغة العربية [Added title page title]
Subject(s): Available additional physical forms:
  • Issued also as CD
Dissertation note: Thesis (M.Sc.) - Cairo University - Institute of Statistical Studies and Research - Department of Computer and Information Science Summary: Multiword expressions (MWEs) refer to any expression composed of two or more words repeated with each other more than one time along text with the same order. MWEs semantically have a meaning that can't be inferred from its candidates; this ambiguity caused a problem to many natural language processing (NLP) application such as tokenization, machine translation, information retrieval, text summarization, etc. as a consequence this ambiguity ends when such applications deal with MWEs as a one unit or word with spaces. Generally, there are three main approaches are used for extracting MEWs statistical approach, linguistic approach, alignment approach, or a combination of two or all of them. We have an assumption which assumes linguistic rules may enhance the obtained results from the statistical phase; so a hybrid approach used to extract Arabic MWEs from three different Arabic corpora; which combines the statistical approach that discover the repeated MWEs and its results enhanced using the linguistic approach. Our method extracted the bigram candidates from Arabic corpus, and it reported a 67% precision compared to previous work results which was 32%, and it evaluated by a human expert
Tags from this library: No tags from this library for this title. Log in to add tags.
Star ratings
    Average rating: 0.0 (0 votes)
Holdings
Item type Current library Home library Call number Copy number Status Date due Barcode
Thesis Thesis قاعة الرسائل الجامعية - الدور الاول المكتبة المركزبة الجديدة - جامعة القاهرة Cai01.18.02.M.Sc.2016.Sa.S (Browse shelf(Opens below)) Not for loan 01010110070486000
CD - Rom CD - Rom مخـــزن الرســائل الجـــامعية - البدروم المكتبة المركزبة الجديدة - جامعة القاهرة Cai01.18.02.M.Sc.2016.Sa.S (Browse shelf(Opens below)) 70486.CD Not for loan 01020110070486000

Thesis (M.Sc.) - Cairo University - Institute of Statistical Studies and Research - Department of Computer and Information Science

Multiword expressions (MWEs) refer to any expression composed of two or more words repeated with each other more than one time along text with the same order. MWEs semantically have a meaning that can't be inferred from its candidates; this ambiguity caused a problem to many natural language processing (NLP) application such as tokenization, machine translation, information retrieval, text summarization, etc. as a consequence this ambiguity ends when such applications deal with MWEs as a one unit or word with spaces. Generally, there are three main approaches are used for extracting MEWs statistical approach, linguistic approach, alignment approach, or a combination of two or all of them. We have an assumption which assumes linguistic rules may enhance the obtained results from the statistical phase; so a hybrid approach used to extract Arabic MWEs from three different Arabic corpora; which combines the statistical approach that discover the repeated MWEs and its results enhanced using the linguistic approach. Our method extracted the bigram candidates from Arabic corpus, and it reported a 67% precision compared to previous work results which was 32%, and it evaluated by a human expert

Issued also as CD

There are no comments on this title.

to post a comment.