ارسل ملاحظاتك

ارسل ملاحظاتك لنا







Morpo-Probabilistic Approch Towards Arabic Non-Words Detection and Correction

المصدر: مجلة التواصل اللساني
الناشر: مؤسسة العرفان للإستشارات التربوية والتطوير المهني
المؤلف الرئيسي: Haddad, Bassam (Author)
مؤلفين آخرين: Hattab, Mamoun (Advisor), Yassen, Mustafa (Advisor)
المجلد/العدد: مج15, ع1,2
محكمة: نعم
الدولة: المغرب
التاريخ الميلادي: 2013
الصفحات: 123 - 141
ISSN: 0851-6774
رقم MD: 596892
نوع المحتوى: بحوث ومقالات
اللغة: الإنجليزية
قواعد المعلومات: AraBase
مواضيع:
رابط المحتوى:
صورة الغلاف QR قانون

عدد مرات التحميل

5

حفظ في:
المستخلص: This paper addresses issues involved in spell checking, non-word recognition and correction of Arabic words. As Arabic is known for its highly inflectional morphological structures, this approach is utilizing morphological knowledge in form of consistent root-pattern relationships, and some morpho-syntactical knowledge based on affixation and morphographemic rules to specify the word recognition and non-word correction process. Furthermore, this paper is proposing some new probabilistic measures for completing the task of the correction by locating, reducing and ranking the most probable correction candidates in derivative words. In this context, based on frequency of occurrence analysis, two probabilistic measures are introduced; Root- Pattern Predictive Value, RPV, and Pattern-Root Predictive Value, PPV. Moreover, keyboard effect as well as letter sound and similarity are considered in the form of conflict sets and lexical features as an additional aid to improve the process of error detection and correction.

ISSN: 0851-6774

عناصر مشابهة